Genetically-Defined Deficiency of Mannose-Binding
Lectin Is Associated with Protection after Experimental
Stroke in Mice and Outcome in Human Stroke
Alvaro Cervera1, Anna M. Planas2, Carles Justicia2, Xabier Urra1, Jens C. Jensenius3, Ferran Torres4,5,
Francisco Lozano5,6*, Angel Chamorro1*
1Comprehensive Stroke Center, Hospital Clı ´nic, Institut d’Investigacions Biome `diques August Pi i Sunyer (IDIBAPS), Medical School, University of Barcelona, Barcelona,
Spain, 2Department of Brain Ischemia and Neurodegeneration, Institut d’Investigacions Biome `diques de Barcelona (IIBB)-Consejo Superior de Investigaciones Cientı ´ficas
(CSIC), Institut d’Investigacions Biome `diques August Pi i Sunyer (IDIBAPS), Barcelona, Spain, 3Department of Medical Microbiology and Immunology, University of Aarhus,
Aarhus, Denmark, 4Clinical Pharmacology Unit, Hospital Clı ´nic, Institut d’Investigacions Biome `diques August Pi i Sunyer (IDIBAPS), Medical School, University of Barcelona,
Barcelona, Spain, 5Department of Cellular Biology, Immunology, and Neuroscience, Medical School, University of Barcelona, Barcelona, Spain, 6Immunology Department,
Hospital Clı ´nic, Institut d’Investigacions Biome `diques August Pi i Sunyer (IDIBAPS), Barcelona, Spain
Background: The complement system is a major effector of innate immunity that has been involved in stroke brain damage.
Complement activation occurs through the classical, alternative and lectin pathways. The latter is initiated by mannose-
binding lectin (MBL) and MBL-associated serine proteases (MASPs). Here we investigated whether the lectin pathway
contributes to stroke outcome in mice and humans.
Methodology/Principal Findings: Focal cerebral ischemia/reperfusion in MBL-null mice induced smaller infarctions, better
functional outcome, and diminished C3 deposition and neutrophil infiltration than in wild-type mice. Accordingly,
reconstitution of MBL-null mice with recombinant human MBL (rhMBL) enhanced brain damage. In order to investigate the
clinical relevance of these experimental observations, a study of MBL2 and MASP-2 gene polymorphism rendering the lectin
pathway dysfunctional was performed in 135 stroke patients. In logistic regression adjusted for age, gender and initial
stroke severity, unfavourable outcome at 3 months was associated with MBL-sufficient genotype (OR 10.85, p=0.008) and
circulating MBL levels (OR 1.29, p=0.04). Individuals carrying MBL-low genotypes (17.8%) had lower C3, C4, and CRP levels,
and the proinflammatory cytokine profile was attenuated versus MBL-sufficient genotypes.
Conclusions/Significance: In conclusion, genetically defined MBL-deficiency is associated with a better outcome after acute
stroke in mice and humans.
Citation: Cervera A, Planas AM, Justicia C, Urra X, Jensenius JC, et al. (2010) Genetically-Defined Deficiency of Mannose-Binding Lectin Is Associated with
Protection after Experimental Stroke in Mice and Outcome in Human Stroke. PLoS ONE 5(2): e8433. doi:10.1371/journal.pone.0008433
Editor: Andreas Meisel, Charite ´ Universitaetsmedizin Berlin, Germany
Received September 25, 2009; Accepted November 26, 2009; Published February 3, 2010
Copyright: ? 2010 Cervera et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Financed in part by the European Community (FP7/2007-2013; grant agreement nu 201024), the Spanish Ministry of Education and Science (SAF2008-
04515-CO2-01 to AMP), the non-profitable foundation ‘Fundacio ´n Melchor Colet’ to A.Ch., and the Spanish Network for the Research in Infectious Diseases (REIPI,
RD06/0008/1013) to F.L. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. A. Cervera
and X. Urra are recipients of grants from the Spanish Institute of Health Carlos III.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: Flozano@clinic.ub.es (FL); email@example.com (AC)
Innate immunity presents the first line of defense against infections
through the recognition of pathogen-associated molecular patterns.
Innate immunity also recognizes molecular motifs expressed by
stressed or dead host cells following a general system of response
of the major effector mechanisms of the innate immunity and its
activation contributes to the release of inflammatory mediators,
opsonophagocytosis and killing of pathogens and removal of damaged
and astrocytes, microglia, neurons, and oligodendrocytes are able to
synthesize several complement components . Recently, some
clinical studies have provided evidence supporting activation of the
complement system in patients with acute stroke [3,4].
Crucial steps of the innate response entail the activation of the
alternative pathway and the lectin pathway of the complement
system, to converge in the generation of enzyme complexes that
are able to cleave C3, which is the most abundant complement
protein . In experimental stroke, the complement system is also
activated through non classical pathways (either the alternative or
the lectin pathways) [6,7]. Complement activation appears to
contribute to the development of larger brain infarctions ,
hence depletion of complement results in beneficial effects [9,10],
more readily observed in situations where the brain ischemia is
followed by reperfusion . However, the mechanism and the
clinical relevance of complement activation in human stroke
Mannose-binding lectin (MBL) is a C-type collectin homologous to
C1q that can initiate the activation of complement through the lectin
PLoS ONE | www.plosone.org1February 2010 | Volume 5 | Issue 2 | e8433
pathway. This is an ancient and highly conserved pathway of
complement activation that has been implicated in the pathophys-
iology of myocardial infarction , gastrointestinal ischemia ,
and kidney ischemia/reperfusion . A recent study shows the
benefits of C1 inhibitor administration in a murine model of cerebral
ischemia/reperfusion and suggests that MBL is involved in this effect
. The activation of the lectin pathway is mediated by associated
serine proteases, termed MASPs, since they were discovered asMBL-
associated serine proteases . Of those, MASP-2 is essential as it
cleaves C4 and C2 components, leading to the formation of the
classical C3 convertase . MBL can also recognize and bind
damaged host cells  and it is possible that this collectin might also
recognize damaged cells in the central nervous system.
Low serum levels of MBL are reported in one in five persons of
European descent due to the presence of single nucleotide
polymorphisms (SNPs) in the promoter and the structural coding
region of the MBL2 gene [17,19]. A SNP in the MASP2 gene
(D105.G) may also result in low circulating levels of MASP-2 
but little is known of its pathogenic consequences after experimental
or clinical stroke. The worldwide expansion of a few MBL2 mutant
alleles that lead to low serum levels of MBL has led to the suggestion
that they might be advantageous under some circumstances . In
this study, we investigated the clinical implications of genetic
polymorphisms of key components of the lectin pathway (MBL2/
MASP2) in patients with acute stroke. We also assessed the
consequences of acute ischemia in MBL-null mice and wild-type
mice. To our knowledge, the results of this study underscore for the
first time the central role of the lectin pathway of complement
illustrate that a richer expression of this soluble pattern recognition
molecule in the innate immune system facilitates a stronger
inflammatory response after acute brain injury.
Materials and Methods
Brain Ischemia in Mice
Four-month-old male MBL-null mice (B6.129S4-Mbl1tm1Kata
Mbl2tm1Kata/J) were obtained from The Jackson’s Laboratory.
Age-matched male C57BL/6J wild-type (WT) mice were used as
controls. Animal work was approved by the local Ethical
Committee (Comite ´ E´tico de Experimentacio ´n Animal, University
of Barcelona). Mice were anesthetized with isoflurane and brain
ischemia was induced by 2-hour occlusion of the middle cerebral
artery (MCA), as reported . Cortical perfusion was evaluated
with laser Doppler flowmetry (Perimed, Sweden). Neurological
deficits were assessed at 48 h. Infarct volume was measured at
48 h in 1 mm-thick coronal brain sections stained with 1%
solution of 2,3,5- triphenyltetrazolium chloride (TTC). The area of
infarction  (pale zone not turning red after staining) was
measured in each section and a correction for oedema was made
by multiplying the infarct area by the ratio of the contralateral to
the ipsilateral hemisphere. The areas were integrated to calculate
infarct volume. In each brain section, we calculated the percentage
of hemispheric area that was infarcted by dividing the infarct area
by the area of the contralateral hemisphere *100 (See Text S1).
Reconstitution with rhMBL
An additional group of MBL-null mice received either
endotoxin-free rhMBL or vehicle (tris-buffered saline, TBS). The
rhMBL was generously supplied by NatImmune A/S, Copenha-
gen. It was produced by transfected HEK293F cells and purified
by carbohydrate affinity chromatography and gel permeation
chromatography. DNA was removed by Benzonase treatment;
and several microfiltration and nanofiltration steps were done to
reduce bioburden and elimination of adventitious virus. No sign of
side effects such as would be expected from endotoxin contam-
ination were seen in pre-clinical or phase I trials . rhMBL
(200 mg in 200 ml of TBS) or vehicle was given i.p. 4 times, i.e.
24 h, 14 h, and 2 h prior to MCA occlusion and 12 h later .
Researchers inducing ischemia and measuring infarct volume
were not aware of the treatment assigned to each animal.
Immunohistochemistry and Western Blotting
Immunohistochemistry was performed in cryostat sections
obtained at 48 h. A rabbit polyclonal antibody against C3
(Abcam, Cambridge, UK) was used diluted 1:500. The percent
immunopositive area for each animal was measured using
AnalySIS Software (Soft Imaging System). Quantification of C3
deposition was made in sections taken at three different coronal
sections form Bregma +2 mm to Bregma 23 mm. Images were
taken from 3 fields and the mean value was calculated for each
animal. The area of each field was 0.32 mm2. Images were taken
from the zones showing the highest intensity of fluorescence per
each section. The sections were labeled with a code that did not
reveal the identity of the animals. Brain protein extracts were
obtained from frozen brain tissue and processed for Western
blotting . Two different rabbit anti-mouse antibodies against
C3 (from Abcam and from Santa Cruz Biotechnology Inc., CA,
USA) were used. Further details are given in Text S1.
Patients were part of the ESPIAS stroke trial . Subjects gave
specific written consent for genetic analyses. The study was
conducted according to the Declaration of Helsinki principles and
approved by an official Internal Review Board (Ethic Committee
of Clinical Investigation of the Hospital Clinic of Barcelona;
number of approval: HCP 01/01237). A modified Rankin Scale
(MRS),2, NIHSS score,2, and Barthel Index of 95 or 100 at 3
months indicated a favourable outcome. Another outcome
measure was the incidence of infections during the first week .
Blood was collected at baseline (day 0) and days 1, 2, 3, 4, 7, and
90, and serum was stored at 280uC. The concentration of
leukocytes, C3 and C4, and of several cytokines and other
molecules was quantified (see Text S1). The balance between T
helper 1 (Th1) and Th2 cytokines was assessed by the relationship
between circulating cytokine levels according to the calculation
Genetic and Functional Assessment of the Lectin
At day 0, genomic DNA was extracted. MBL2 and MASP2 were
genotyped with reported sequence-based typing technique 
(see Text S1). According to previous studies [19,28,29], genotypes
0/0, 0/XA, and XA/XA were classified as MBL-low variants
while the remainders were MBL-sufficient variants. For MASP2
gene, SNP at codon 105 exon 3 implies a non-synonymous amino
acid replacement (D105.G) at the CUB1 domain of MASP-2,
which impedes association to MBL and significantly reduces
Serum concentrations (ng/ml) of MBL and MASP-2 were
(TRIFMA) [30,31] (see Text S1).
Categorical data were compared using the x2and Fisher’s exact
tests. Continuous variables were analyzed with the Student’s t-test,
non-parametric Mann-Whitney U-test, and Pearson or Spearman
MBL in Acute Stroke
PLoS ONE | www.plosone.org2 February 2010 | Volume 5 | Issue 2 | e8433
coefficients when required. ANCOVA adjusted for baseline values
was used to assess serial changes of inflammatory parameters
according to genotype. Two-way ANOVA was applied for
comparisons of the percentage of infarcted area between MBL-
null mice and WT mice for the different brain sections. Changes
over time of MBL and MASP-2 levels were analysed with a
longitudinal mixed model for repeated measurements that
included genotype, time and their interaction in the model, and
setting the type of covariance as unstructured. Logistic regression
modeling was used to assess baseline predictors of functional
independence at day 90 adjusted for the effects of age, gender and
baseline NIHSS score.
MBL-Deficiency Is Beneficial in Experimental Brain
Brain damage after 2-hour middle cerebral artery occlusion
(MCA) was assessed by measuring infarct volume at 48 h in WT
(n=10) and MBL-null (n=11) mice. Representative images of
infarction in WT and MBL-null mice are shown in Fig. 1A. Infarct
volume was significantly reduced (p,0.05) in the MBL-null group
(Fig. 1B). The effects were significant in cortical (p,0.05) (Fig. 1C)
and subcortical (p,0.01) (Fig. 1D) structures. However, as the
mean6SD body weight of MBL-null mice (26.5060.46 g) was
smaller than that of WT mice (29.0560.93 g) (p,0.001), we made
some calculations to reassure that the reduction in body size in the
MBL-null mice did not contribute to the differences in infarct
volume vs. WT. In each brain section, we calculated the
percentage of hemispheric area that was infarcted by dividing
the infarct area by the area of the contralateral hemisphere *100.
The results (Fig. 1E) showed that the proportion of brain tissue
that was infarcted in the different brain sections was smaller in
MBL-null mice than in WT (p,0.001, two-way ANOVA by
genotype and brain section). In agreement with the reduction in
infarct size, the neurological test studied at 48 h showed that the
functional deficit was minor in MBL-null mice than in WT mice
(p=0.05) (Fig. 1F).
The magnitude of the ischemic lesion is dependent on the
reduction of cerebral blood flow (CBF) experimentally induced by
occlusion of the middle cerebral artery (MCA). For this reason,
we verified that MBL-deficient mice responded to MCA
occlusion with a similar perfusion deficit than WT mice by
monitoring CBF with laser Doppler flowmetry. The results
showed that CBF (mean6SD) dropped until 3068% and
3267% of baseline in WT mice (n=6) and in MBL-null mice
(n=8), respectively. The results illustrated that MBL-deficiency
did not affect (p=0.61) the CBF response to MCA occlusion and,
thus, that the severity of the induced ischemia was similar in WT
and MBL-null mice.
We then carried reconstitution experiments by administering
human recombinant MBL (rhMBL) protein to MBL-null mice to
validate that the smaller infarct volume in these animals was
attributable to the lack of MBL. rhMBL or vehicle was
administered i.p. following a reconstitution protocol that rescues
MBL pathway activity in plasma . MBL-null mice that were
reconstituted with rhMBL showed a significantly larger infarction
than mice of the vehicle group (p,0.05) (Fig. 1G), in spite that a
similar CBF reduction after ischemia was observed after rhMBL or
vehicle (p=0.84). Also, a trend was observed for a worst
neurological score in the rhMBL group, but the difference
compared with vehicle did not reach statistic significance
(p=0.09) (Fig. 1H).
Post-Ischemic Complement Activation Was Attenuated in
MBL-Null Mice vs. WT
We examined the deposition of C3 complement protein in
mouse brain by means of immunohistochemistry in frozen sections
of brain tissue obtained at 48 h post-ischemia. C3 immunoreac-
tivity was observed in blood vessels and also within the brain
parenchyma in the ipsilateral hemisphere of WT mice, and, to a
lesser extent, of MBL-null mice (Fig. 2A). Counterstaining the C3
immunoreaction with Hoechst showed C3 positive reaction
surrounding certain cell bodies (Fig. 2B). For a quantitative
assessment of differences in C3 immunoreactivity between groups
we measured the C3-positive area in microscope images taken
from the ischemic zone. The results (Fig. 2C) showed that the C3-
immunoreactive area was smaller in MBL-null (n=5) than in WT
(n=5) mice, thus supporting that the extent of complement
activation was milder in MBL-null than in WT mice.
Native C3 is a heterodimeric protein made of one a-chain
(115-kDa) and one b-chain. C3 activation causes cleavage of C3a
and generationofanactive C3afragment (9 kDa) andtheremaining
fragment named C3ba’. The latter can covalently bind to small
target proteins and acquire a greater apparent molecular mass.
Evidence for C3 cleavage after brain ischemia was obtained by
studying frozen brain tissue that was homogenized with detergent
and examined by Western blotting using development with anti-C3
antibody. In the ischemic brain tissue of WT mice, the full length C3
protein and the C3a chain were degraded, as shown in Fig. 3A,
suggesting cleavage of the protein. This effect occurred to a lesser
extent in ischemic MBL-null mice (Fig. 3A and B). Moreover, we
detected a C3 fragment of around 120 kDa in the ischemic brain
tissue of WT mice, which was absent in the control brain, and less
abundant in MBL-null than in WT mice (Fig. 3C and D). This
protein had a molecular weight compatible with C3ba’.
As an indicator of neutrophil accumulation in the ischemic
brain, expression of myeloperoxidase (MPO) was assessed by
Western blotting in frozen brain tissue using an anti-MPO
antibody. Ischemia induced MPO accumulation in the brain of
WT mice, and this effect was milder in MBL-null mice (Fig. 3E
and F), suggesting that neutrophil accumulation after ischemia was
attenuated in the brain of MBL-deficient animals.
The Mannose-Pathway in Human Stroke
We undertook investigations to find out whether the neuropro-
tective effects of MBL-deficiency found in experimental brain
ischemia are relevant to human stroke. For this purpose, we studied
the genotype, and measured MBL levels, together with C3 and C4
complement proteins, and several markers of inflammation in the
serum of our patients. One-hundred and thirty five patients
completed the study, including 109 (80.7%) ischemic strokes and
26 (19.3%) hemorrhagic strokes. Patients with ischemic or
hemorrhagic stroke disclosed no significant differences in demo-
graphics, risk factors, stroke severity on admission, clinical course,
MBL and MAPS2 genotype and levels (see Table S1). Infections
during the first 7 days after stroke were observed in 24 (17.8%)
patients, functional independence at day 90 was reached by 47
(34.8%) patients, and there were 25 deaths at day 90 (19%). Other
general traits of these patients were as previously reported .
In our study, twenty-four (17.8%) patients had genetic MBL-low
variants. These patients showed lower serum levels of MBL
protein than the group with genetic MBL-sufficient variants
(Fig. 4A). MBL levels in serum disclosed no major fluctuations
over time after stroke onset (p=0.79). In addition, fifteen (11.1%)
patients were heterozygous for the D105.G SNP at the MASP2
gene, while none of the patients were homozygous. This SNP has
been linked to a severe immunodeficiency when presented in
MBL in Acute Stroke
PLoS ONE | www.plosone.org3 February 2010 | Volume 5 | Issue 2 | e8433
Figure 1. MBL-null mice develop smaller infarct volumes than wild-type mice. A) Representative images of infarcted brain tissue in wild-
type (WT) and MBL-null mice at 48 h post-ischemia. The approximate level of the brain sections is indicated in the left by the distance from Bregma.
B) Infarct volume was lesser in the MBL-null group (n=11) than in WT (n=10). C) Cortical and D) subcortical infarct volumes were smaller in MBL-null
than in WT mice. E) The percentage of tissue with infarction per brain section (1 mm-thick) is shown for consecutive sections, starting (left) from the
frontal part of the brain. The proportion of infarcted tissue is smaller in MBL-null than in WT mice. Values are expressed as the mean6SD. Two-way
ANOVA by genotype and brain section showed significant differences due to genotype (p,0.001, F(9,40)=18.74). Post-hoc Bonferroni test showed
significant differences in the indicated brain sections. F) The neurological score was better (lower) in MBL-null mice than in WT. (G–H) MBL-null mice
were treated with rhMBL (MBLnull+rhMBL) (n=4) or vehicle (MBLnull) (n=5). G) MBL-null mice receiving rhMBL showed larger infarct volume than
non-reconstituted mice. H) A trend to worst neurological deficit (p=0.09) was seen after reconstitution with rhMBL compared to vehicle. Symbols
indicate values for individual animals.*p,0.05, ** p,0.01, *** p,0.001).
MBL in Acute Stroke
PLoS ONE | www.plosone.org4 February 2010 | Volume 5 | Issue 2 | e8433
Figure 2. Ischemia-induced C3 deposition in brain parenchyma is attenuated in MBL-null versus WT mice. A) C3 deposition (red),
laminin immunostaining (blue), and CD11b (green). C3 immunoreaction is often associated to vessels and it is also found within the ischemic brain
parenchyma. The C3 reaction is more moderate in MBL-null mice than in the WT. Control indicates non-operated WT mice. Bar scale =10 mm. B) C3
deposition (red) and Hoechst counterstaining to illustrate the cell nuclei shows C3 immunoreaction surrounding certain cell bodies (arrow in b). (b) is
a magnification of the square shown in (a). C) Quantification of the C3 immunoreactive area in the ipsilateral hemisphere shows a significant
reduction in MBL-null mice (n=5) versus the WT (n=5). Values are expressed as the mean6SD. Bar scale: (a) 100 mm, (b) 10 mm. * indicates p,0.05.
Figure 3. C3 activation and neutrophil infiltration in the ischemic brain is lower in MBL-null than WT mice. A) C3 a-chain is observed in
control tissue with a rabbit anti-mouse C3 antibody. C3 is cleaved after ischemia in WT mice and, to a lesser extent, in MBL-null mice. Lanes represent
different animals. B) Quantification of band optical density shows more severe cleavage of the C3 a-chain in WT than in MBL-null mice. C) The C3ba’
fragment (120 kDa) is evidenced in the ischemic tissue with an anti-C3 antibody. D) Quantification of band optical density shows that C3ba’ is more
abundant in the ischemic tissue of WT than of MBL-null mice. E) Myeloperoxidase (MPO) is a marker of neutrophil infiltration in brain tissue. MPO is
detected in the ischemic brain of WT mice, while it is attenuated in MBL-KO mice, as shown in F). Values are expressed as mean6SD. * p,0.05.
MBL in Acute Stroke
PLoS ONE | www.plosone.org5 February 2010 | Volume 5 | Issue 2 | e8433
homozygousity [20,32]. Patients with this MASP2 polymorphism
disclosed lower levels of MASP-2 protein (Fig. 4B). Overall, there
were no significant changes in circulating levels of MASP-2 over
time (p=0.82). Demographics, risk factors, clinical parameters at
admission, and treatment allocation were evenly distributed
according to MBL2 or MASP2 genotypes, as shown in Table 1.
Figure 4. Serum levels of MBL and MASP-2 and complement system activation according to MBL2 and MASP2 genotypes. A) Serum
concentrations (ng/ml) of MBL in MBL-low genotypes and MBL-sufficient genotypes. B) Serum concentrations (ng/ml) of MASP-2 in patients with
D105.G or wild-type MASP2 genotypes. Measurements were performed at day 0 (d0) (n=96), day 1 (d1) (n=10), day 2 (d2) (n=7), day 3 (d3) (n=10),
day 4 (d4) (n=10), day 7 (d7) (n=9) and day 90 (d90) (n=96). C) Serum concentration (g/L) of C3 and C4 at day 0 (d0) (n=96) and day 2 (d2) (n=96).
Values are the mean 6SD. *p,0.05; **p,0.01; ***p,0.0001.
Table 1. Baseline characteristics in the study population (n=135) according to MBL and MASP2 genotype.
D105. .G MASP2-WTp VALUE
24 (18)111 (82) 15 (11)120 (89)
AGE (MEAN, SD), YRS
74 (13)73 (12)0.67 70 (15) 73 (11)0.48
MALE, NO.(%) 9 (38) 59 (53)0.16 7 (47)61 (51)0.49
ACTIVE SMOKING, NO. (%)3 (13) 20 (18)0.77 3 (20) 20 (17)0.39
HYPERTENSION, NO. (%) 12 (50)73 (66)0.156 (40)79 (66) 0.06
DIABETES, NO. (%) 6 (25)24 (22)0.72 5 (33) 25 (21)0.27
CORONARY HEART DISEASE, NO. (%) 2 (8) 15 (14)0.49 1 (7)16 (13)0.22
PREVIOUS STROKE, NO. (%)4 (17) 19 (17) 1.000 (0) 23 (19)0.06
PERIPHERAL ARTERY DISEASE, NO. (%)2 (8)9 (8)1.000 (0)11 (9)0.22
ADMISSION NIHSS SCORE, MEAN. (SD)12 (7) 14 (7) 0.1014 (7)13 (7)0.60
HEMORRHAGIC STROKE AT ONSET, NO. (%) 6 (25) 20 (18)0.403 (20)23 (19)0.93
SYSTOLIC BP (MEAN, SD), MM HG
165 (39)163 (30)0.80 160 (34)164 (32) 0.69
DIASTOLIC BP (MEAN, SD), MM HG
87 (21) 89 (20) 0.7194 (24)88 (19) 0.52
GLUCOSE (MEAN, SD), MG/DL141 (36)145 (54)0.69166 (70)141 (47) 0.19
AXILLARY TEMPERATURE (MEAN, SD), uC36 (0.5)36 (0.5)0.3236 (0.5) 36 (0.5)0.35
TREATMENT WITH LEVOFLOXACIN, NO. (%) 12 (50)55 (50)0.97 10 (67)57 (48)0.16
MBL in Acute Stroke
PLoS ONE | www.plosone.org6February 2010 | Volume 5 | Issue 2 | e8433
Patients with MBL-low genotypes disclosed lower serum levels
of C3 and C4 than patients with MBL-sufficient genotypes
(Fig. 4C). Likewise, MBL levels were positively correlated with C3
at different time points after stroke onset, i.e. day 0 (r=0.23,
p=0.01) and at day 2 (r=0.34, p,0.01), and C4 at day 0
(r=0.31, p,0.01), and at day 2 (r=0.36, p=0.001). In contrast,
the MASP2 genotype did not influence the levels of C3 and C4
after stroke (Fig. S1).
Circulating C-Reactive Protein and Cytokine Profile of
Stroke Patients According to Their Genotype
CRP levels were lower in patients with MBL-low genotypes
compared to MBL-sufficient genotype (Fig. 5A), but were not
related with the MASP2 genotype (data not shown). In exploratory
analyses restricted to patients without infections, a lower rise of
CRP was confirmed in patients with MBL-low variants (Fig. 5B).
Expectedly, CRP and IL-6 were highly correlated (r=0.37,
p,0.0001). CRP was also positively correlated with C3 (r=0.44,
p,0.0001), and C4 (r=0.33, p,0.01).
MBL-low patients showed a cytokine profile after stroke
characterized by a greater increase of IL-10, a lower rise of IL-
6, and no significant changes of TNF-a compared to patients with
MBL-sufficient genotypes (Fig. 6A–D). This predominant anti-
inflammatory cytokine response remained significant 3 months
after stroke onset. The temporal course of other immune players,
such as neutrophils, monocytes, and lymphocytes, did not show a
significant effect related to MASP2 or MBL2 genotypes (data not
Stroke Recovery in Patients
Functional independence at day 90 was observed in 14 patients
(58.3%) with MBL-low, and in 33 patients (29.7%) with MBL-
sufficient genotype (x2, p=0.008). Unfavourable outcome at 3
months after stroke was associated with MBL-sufficient genotype
(OR 10.85, 95% CI 62.94–1.87, p=0.008) and MBL levels (OR
1.29, 95% CI 1.69–1.02, for every 500 units increase, p=0.04) in
a logistic regression model adjusted for age (OR 1.06, 95% CI
1.00–1.12, p=0.04), sex (OR 0.32, 95% CI 0.09–1.16, versus
male gender, p=0.08), and initial stroke severity (OR 4.14, 95%
CI 2.20–7.88, for each quartile of NIHSS score, p,0.0001). The
MASP2-sufficient genotype (OR 0.92, 95% CI 0.23–3.73,
p=0.91) was not associated with stroke recovery. And last,
patients with unfavourable outcome showed a predominantly pro-
inflammatory cytokine profile (Fig. S2).
Post-stroke infections at day 7 were observed in 5 (20.8%)
patients with MBL-low genotypes and 19 (17.1%) patients with
MBL-sufficient genotypes (x2, p=0.67). Regarding the MASP2
genotype, post-stroke infections appeared in 2 (13.3%) patients
with D105.G SNP and in 22 (18.3%) patients with the wild-type
genotype (x2, p=0.63). The risk of infection after stroke was not
associated with the serum levels of MBL at day 0 (p=0.47) or at
day 90 (p=0.95), the levels of MASP-2 at day 0 (p=0.07) or at
day 90 (p=0.14), the MBL genotype (OR=1.27, 95% CI 0.42–
3.84, p=0.67), or the MASP2 genotype (OR=0.68, 95% CI 0.14–
The complement is an integral component of the innate
immune system that has been implicated in the pathophysiology of
acute stroke [3,4,33]. In this study, we report for the first time the
important contribution of the lectin pathway of complement
activation to brain tissue fate after acute stroke, both in mice and
humans. All the patients included in this study participated in a
randomized, double-blind, controlled trial that limited the
likelihood of clinical bias, and the main clinical findings of the
study were in consonance with the results in experiments
conducted in WT and MBL-null mice.
MBL-null mice disclosed after stroke smaller infarctions, and
better functional outcome than WT mice. This effect was
associated with attenuation of complement deposition and
neutrophil accumulation into the ischemic brain tissue in
MBL-null versus WT mice. The contribution of MBL to
ischemic brain damage was further confirmed when MBL was
exogenously reconstituted in MBL-null mice. In agreement with
these findings in experimental animals, the main clinical
contribution of this study was that MBL2 genotypes, which were
associated with low circulating levels of functional MBL protein
oligomers, increased approximately 11 times the odds of good
functional outcome after stroke. Indeed, carriers of this genetic
trait doubled the chance of becoming functionally independent
stroke survivors, independently of the effects of established key
prognostic factors such as age, gender, and initial stroke severity.
Figure 5. C-reactive protein levels (mg/dL) in MBL-low
genotypes and MBL-sufficient genotypes. A) All patients, day 0
(d0) baseline (n=127; MBL-low=23, MBL-sufficient=104); day 1 (d1)
(n=131; MBL-low=24, MBL-sufficient=107); day 2 (d2) (n=132; MBL-
low=24, MBL-sufficient=108); day 3 (d3) (n=127; MBL-low=24, MBL-
sufficient=103); day 4 (d4) (n=122; MBL-low=22, MBL-sufficient=100),
day 7 (d7) (n=111; MBL-low=19, MBL-sufficient=92), and day 90 (d90)
(n=68; MBL-low=12, MBL-sufficient=56). B) Patients without post-
stroke infections, d0 (n=103; MBL-low=18, MBL-sufficient=85); d1
(n=107; MBL-low=19, MBL-sufficient=88); d2 (n=108; MBL-low=19,
MBL-sufficient=89); d3 (n=105; MBL-low=19, MBL-sufficient=86); d4
(n=101; MBL-low=17, MBL-sufficient=84), d7 (n=91; MBL-low=15,
MBL-sufficient=76), and d90 (n=60; MBL-low=11, MBL-sufficient=49).
Values were obtained from serum and are expressed as the mean 6 SD.
* p,0.05, ** p=0.01. *p,0.05, **p=0.01.
MBL in Acute Stroke
PLoS ONE | www.plosone.org7 February 2010 | Volume 5 | Issue 2 | e8433
Importantly, the benefits of a MBL-low genotype on stroke
recovery were not counterbalanced by an increased risk of
infections which have been recently associated with a stroke-
induced immune-depression syndrome . These results
concur with a recent study suggesting that MBL is involved in
the protective effect of C1 inhibitor in a murine model of brain
Unlike the major contribution of MBL, the study did not
unravel a significant involvement of MASP2 genotype or MASP-2
levels in the prognosis of stroke. It has been speculated that MASP-
2 deficiency might have broader consequences than MBL
deficiency because MASP-2 is involved in the biological activity
of not only MBL but also ficolins, a group of functionally related
proteins . However, we did not find an association between
heterozygous MASP2 variant genotypes and stroke outcome. Since
heterozygosity for the MASP2 variant allotypes does not lead to
significant reduction in the function of the lectin pathway, it is
likely that the D105.G variant is only relevant in homozygousity,
which, with a gene frequency of about 5%, is a rare condition .
Nonetheless, on the basis of our findings we cannot fully exclude
the possibility that MBL might have effects independent of MASP-
2, as suggested by several findings. For instance, MBL stimulates
macrophage phagocytosis in a manner independent of MASP-2
. This matter deserves further investigation.
The inflammatory and lytic effects of complement may
significantly contribute to tissue damage and MBL may modulate
inflammation and trigger pro-inflammatory cytokine release from
monocytes . Although we found no significant associations
between the MBL2 genotype and the total number of circulating
monocytes or other white blood cells, we did not assess the
phenotype of monocytes in this cohort. However, recent studies by
our group have shown assorted effects of stroke on circulating
monocytes including an early rise of minor subpopulations with
less inflammatory drive and increased capacity for tissue repair
. Therefore, further investigation will be required to identify
the phenotypic traits of monocytes of subjects with MBL-low
The study identified that the individuals carrying MBL-low
genotypes had lower peaks of CRP, C3 and C4 in serum, and a
strong correlation between CRP, C3 and C4 complement
proteins, in agreement with experiments indicating that the
complement-activating functions of CRP and MBL are coordi-
nated . CRP is a classic acute phase reactant that belongs to
the pentraxin family of calcium dependent ligand-binding plasma
proteins that play a role in the humoral innate response. CRP
binds with high affinity to phosphocholine residues and with a
variety of autologous and extrinsic ligands which are more
extensively exposed in or on damaged cells and tissues (reviewed
in ). Animal studies have shown that CRP can enhance
ischemic tissue damage by a complement-dependent mechanism
in the heart  and brain . The results of our study indicate
that the deleterious effects of MBL after acute stroke are associated
with increased concentrations of CRP, although additional studies
will be required to determine whether there is a causal effect.
Patients with MBL-low genotype also disclosed a cytokine profile
in blood characterized by a predominance of anti-inflammatory
Figure 6. Cytokine levels (pg/ml) in blood in MBL-low genotypes and MBL-sufficient genotypes. A) Interleukin-10; B) Interleukin-6; C)
TNF-a; D) Balance between T helper (h) 1 cytokines and Th2 cytokines calculated as (TNF-a + IL-6)/IL-10. Measurements were performed in serum
samples at day 0 (d0) (n=129; MBL-low=23, MBL-sufficient=106); day1 (d1) (n=121; MBL-low=22, MBL-sufficient=99); day 2 (d2) (n=122; MBL-
low=22, MBL-sufficient=100); day 3 (d3) (n=116; MBL-low=19, MBL-sufficient=97); day 4 (d4) (n=111; MBL-low=16, MBL-sufficient=95), day 7
(d7) (n=101; MBL-low=16, MBL-sufficient=85), and day 90 (d90) (n=92; MBL-low=15, MBL-sufficient=77). Values are the mean 6 SD. *p,0.05,
MBL in Acute Stroke
PLoS ONE | www.plosone.org8 February 2010 | Volume 5 | Issue 2 | e8433
over pro-inflammatory cytokines. Furthermore, the persistence of
this anti-inflammatory profile at 3 months after stroke was in
stronger support of a genetic trait rather than an acute-phase
reaction. It can be argued that the improved functional recovery
and increased brain tissue survival associated with a MBL-low
genotype were related in part with a cytokine environment that
allowed brain cells to be exposed to stronger anti-inflammatory
In summary, this study shows the major relevance of the MBL
pathway of complement activation in acute ischemic or
hemorrhagic stroke. Although the mechanisms that lead to
ischemic and hemorrhagic stroke may differ, our results are in
agreement with the concept of a danger model in which signals of
cell injury sensed by immune receptors trigger shared immune
responses . Our data indicate that a genetically-defined MBL
deficiency facilitates anti-inflammatory responses after acute
stroke that result in long lasting beneficial effects on post-stroke
functional recovery. Validation of these results would justify the
quest of therapeutic interventions aimed to specifically inhibit the
lectin pathway of complement activation in patients with acute
Found at: doi:10.1371/journal.pone.0008433.s001 (0.04 MB
D105.G and WT MASP2-genotypes. Serum concentration (g/
L) of C3 and C4 at day 0 (d0) (n=96) and day 2 (d2) (n=96).
Values are represented as meanAˆ6SD.
Found at: doi:10.1371/journal.pone.0008433.s002 (2.05 MB TIF)
Complement system activation in carriersof
cytokines and clinical outcome. Serum measurements at day 0 (d0)
(n=129); day 1 (d1) (n=121); day 2 (d2) (n=122); day 3 (d3)
(n=116); day 4 (d4) (n=111), day 7 (d7) (n=101), and day 90
(d90) (n=92). Values are mean Aˆ6 SD. *p,0.05.
Found at: doi:10.1371/journal.pone.0008433.s003 (0.67 MB TIF)
Balance between T helper (h) 1 cytokines and Th2
Found at: doi:10.1371/journal.pone.0008433.s004 (0.03 MB
Main traits in patients with ischemic or hemorrhagic
We thank Ms. Bele ´n Sua ´rez for helpful technical assistance.
Conceived and designed the experiments: AC AMP FL AC. Performed the
experiments: AC AMP CJ XU JCJ. Analyzed the data: AC AMP FT FL
AC. Contributed reagents/materials/analysis tools: JCJ. Wrote the paper:
AC AMP AC.
1. Carroll MC, Holers VM (2005) Innate autoimmunity. Adv Immunol 86: 37–57.
2. van Beek J, Elward K, Gasque P (2003) Activation of complement in the central
nervous system: roles in neurodegeneration and neuroprotection. Ann N Y Acad
Sci 992: 56–71.
3. Mocco J, Wilson DA, Komotar RJ, Sughrue ME, Coates K, et al. (2006)
Alterations in plasma complement levels after human ischemic stroke.
Neurosurgery 59: 28–33.
4. Lindsberg PJ, Ohman J, Lehto T, Karjalainen-Lindsberg ML, Paetau A, et al.
(1996) Complement activation in the central nervous system following blood-
brain barrier damage in man. Ann Neurol 40: 587–596.
5. Frank MM, Fries LF (1991) The role of complement in inflammation and
phagocytosis. Immunol Today 12: 322–326.
6. Mocco J, Mack WJ, Ducruet AF, Sosunov SA, Sughrue ME, et al. (2006)
Complement component C3 mediates inflammatory injury following focal
cerebral ischemia. Circ Res 99: 209–217.
7. De Simoni MG, Rossi E, Storini C, Pizzimenti S, Echart C, et al. (2004) The
powerful neuroprotective action of C1-inhibitor on brain ischemia-reperfusion
injury does not require C1q. Am J Pathol 164: 1857–1863.
8. Atkinson C, Zhu H, Qiao F, Varela JC, Yu J, et al. (2006) Complement-
dependent P-selectin expression and injury following ischemic stroke. J Immunol
9. Huang J, Kim LJ, Mealey R, Marsh HC, Zhang Y, et al. (1999) Neuronal
protection in stroke by an sLex-glycosylated complement inhibitory protein.
Science 285: 595–599.
10. Cowell RM, Plane JM, Silverstein FS (2003) Complement activation contributes
to hypoxic-ischemic brain injury in neonatal rats. J Neurosci 23: 9459–9468.
11. Ducruet AF, Hassid BG, Mack WJ, Sosunov SA, Otten ML, et al. (2008) C3a
receptor modulation of granulocyte infiltration after murine focal cerebral
ischemia is reperfusion dependent. J Cereb Blood Flow Metab 28: 1048–1058.
12. Jordan JE, Montalto MC, Stahl GL (2001) Inhibition of mannose-binding lectin
reduces postischemic myocardial reperfusion injury. Circulation 104: 1413–1418.
13. Hart ML, Ceonzo KA, Shaffer LA, Takahashi K, Rother RP, et al. (2005)
Gastrointestinal ischemia-reperfusion injury is lectin complement pathway
dependent without involving C1q. J Immunol 174: 6373–6380.
14. de Vries B, Walter SJ, Peutz-Kootstra CJ, Wolfs TG, van Heurn LW, et al.
(2004) The mannose-binding lectin-pathway is involved in complement
activation in the course of renal ischemia-reperfusion injury. Am J Pathol 165:
15. Gesuete R, Storini C, Fantin A, Stravalaci M, Zanier E, et al. (2009)
Recombinant C1-inhibitor in Brain Ischemic Injury. Ann Neurol. In press.
16. Schwaeble W, Dahl MR, Thiel S, Stover C, Jensenius JC (2002) The mannan-
binding lectin-associated serine proteases (MASPs) and MAp19: four compo-
nents of the lectin pathway activation complex encoded by two genes.
Immunobiology 205: 455–466.
17. Dommett RM, Klein N, Turner MW (2006) Mannose-binding lectin in innate
immunity: past, present and future. Tissue Antigens 68: 193–209.
18. Ogden CA, deCathelineau A, Hoffmann PR, Bratton D, Ghebrehiwet B, et al.
(2001) C1q and mannose binding lectin engagement of cell surface calreticulin
and CD91 initiates macropinocytosis and uptake of apoptotic cells. J Exp Med
19. Madsen HO, Garred P, Thiel S, Kurtzhals JA, Lamm LU, et al. (1995) Interplay
between promoter and structural gene variants control basal serum level of
mannan-binding protein. J Immunol 155: 3013–3020.
20. Stengaard-Pedersen K, Thiel S, Gadjeva M, Moller-Kristensen M, Sorensen R,
et al. (2003) Inherited deficiency of mannan-binding lectin-associated serine
protease 2. N Eng J Med 349: 554–560.
21. Casanova JL, Abel L (2004) Human Mannose-binding Lectin in Immunity:
Friend, Foe, or Both? J Exp Med 199: 1295–1299.
22. Justicia C, Martı ´n A, Rojas S, Gironella M, Cervera A, et al. (2006) Anti-
VCAM-1 antibodies did not protect against ischemic damage neither in rats nor
in mice. J Cereb Blood Flow Metab 26: 321–322.
23. Clark WM, Lessov NS, Dixon MP, Eckenstein F (1997) Monofilament middle
cerebral artery occlusion in the mouse. Neurol Res 19: 642–648.
24. Petersen KA, Matthiesen F, Agger T, Kongerslev L, Thiel S, Cornelissen K,
Axelsen M (2006) Phase I safety, tolerability, and pharmacokinetic study of
recombinant human mannan-binding lectin. J Clin Immunol 26: 465–475.
25. Møller-Kristensen M, Wang W, Ruseva M, Thiel S, Nielsen S, et al. (2005)
Mannan-binding lectin recognizes structures on ischemic reperfused mouse
kidneys and is implicated in tissue injury. Scand J Immunol 61: 426–434.
26. Chamorro A, Horcajada JP, Obach V, Vargas M, Revilla M, et al. (2005) The
Early Systemic Prophylaxis of Infection after Stroke study: a randomized clinical
trial. Stroke 36: 1495–1500.
27. Lozano F, Sua ´rez B, Mun ˜oz A, Jensenius JC, Mensa J, et al. (2005) Novel
MASP2 variants detected among North African and Sub-Saharan individuals.
Tissue Antigens 66: 131–135.
28. Turner MW, Hamvas RM (2000) Mannose-binding lectin: structure, function,
genetics and disease associations. Rev Immunogenet 2: 305–322.
29. Mullighan CG, Heatley S, Doherty K, Szabo F, Grigg A, et al. (2002) Mannose-
binding lectin gene polymorphisms are associated with major infection following
allogeneic hemopoietic stem cell transplantation. Blood 99: 3524–3529.
30. Thiel S, Møller-Kristensen M, Jensen L, Jensenius JC (2002) Assays for the
functional activity of the mannan-binding lectin pathway of complement
activation. Immunobiology 205: 446–454.
31. Møller-Kristensen M, Jensenius JC, Jensen L, Thielens N, Rossi V, et al. (2003)
Levels of mannan-binding lectin-associated serine protease-2 in healthy
individuals. J Immunol Methods 282: 159–167.
32. Sorensen R, Thiel S, Jensenius JC (2005) Mannan-binding-lectin-associated
serine proteases, characteristics and disease associations. Springer Semin
Immunopathol 27: 299–319.
33. Pedersen ED, Waje-Andreassen U, Vedeler CA, Aamodt G, Mollnes TE (2004)
Systemic complement activation following human acute ischaemic stroke. Clin
Exp Immunol 137: 117–122.
MBL in Acute Stroke
PLoS ONE | www.plosone.org9February 2010 | Volume 5 | Issue 2 | e8433
34. Chamorro A, Urra X, Planas AM (2007) Infection after acute ischemic stroke: a
manifestation of brain-induced immunodepression. Stroke 38: 1097–1103.
35. Shiratsuchi A, Watanabe I, Ju JS, Lee BL, Nakanishi Y (2008) Bridging effects of
recombinant human mannose-binding lectin in macrophage phagocytosis of
Escherichia coli. Immunology 124: 575–583.
36. Chaka W, Verheul AF, Vaishnav VV, Cherniak R, Scharringa J, et al. (1997)
Induction of TNF-alpha in human peripheral blood mononuclear cells by the
mannoprotein of Cryptococcus neoformans involves human mannose binding
protein. J Immunol 159: 2979–2985.
37. Urra X, Villamor N, Amaro S, Go ´mez-Choco M, Obach V, et al. (2009)
Monocyte subtypes predict clinical course and prognosis in human stroke.
J Cereb Blood Flow Metab 29: 994–1002.
38. Kaplan MH, Volanakis JE (1974) Interaction of C-reactive protein complexes
with the complement system. I. Consumption of human complement associated
with the reaction of C-reactive protein with pneumococcal C-polysaccharide and
with the choline phosphatides, lecithin and sphingomyelin. J Immunol 112:
39. Di Napoli M, Schwaninger M, Cappelli R, Ceccarelli E, Di Gianfilippo G, et al.
(2005) Evaluation of C-reactive protein measurement for assessing the risk and
prognosis in ischemic stroke: a statement for health care professionals from the
CRP Pooling Project members. Stroke 36: 1316–1329.
40. Griselli M, Herbert J, Hutchinson WL, Taylor KM, Sohail M, et al. (1999) C-
reactive protein and complement are important mediators of tissue damage in
acute myocardial infarction. J Exp Med 190: 1733–1740.
41. Gill R, Kemp JA, Sabin C, Pepys MB (2004) Human C-reactive protein
increases cerebral infarct size after middle cerebral artery occlusion in adult rats.
J Cereb Blood Flow Metab 24: 1214–1218.
42. Planas AM, Chamorro A (2009) Regulatory T cells protect the brain after stroke.
Nat Med 15: 138–139.
43. Matzinger P (2002) The danger model: a renewed sense of self. Science 296:
MBL in Acute Stroke
PLoS ONE | www.plosone.org10 February 2010 | Volume 5 | Issue 2 | e8433