Article
Reference
The junctional adhesion molecule JAM-C regulates polarized
transendothelial migration of neutrophils in vivo
WOODFIN, Abigail, et al.
Abstract
The migration of neutrophils into inflamed tissues is a fundamental component of innate
immunity. A decisive step in this process is the polarized migration of blood neutrophils
through endothelial cells (ECs) lining the venular lumen (transendothelial migration (TEM)) in
a luminal-to-abluminal direction. By real-time confocal imaging, we found that neutrophils had
disrupted polarized TEM ('hesitant' and 'reverse') in vivo. We noted these events in
inflammation after ischemia-reperfusion injury, characterized by lower expression of junctional
adhesion molecule C (JAM-C) at EC junctions, and they were enhanced by blockade or
genetic deletion of JAM-C in ECs. Our results identify JAM-C as a key regulator of polarized
neutrophil TEM in vivo and suggest that reverse TEM of neutrophils can contribute to the
dissemination of systemic inflammation.
WOODFIN, Abigail, et al. The junctional adhesion molecule JAM-C regulates polarized
transendothelial migration of neutrophils in vivo. Nature immunology, 2011, vol. 12, no. 8, p.
761-9
DOI : 10.1038/ni.2062
PMID : 21706006
Available at:
http://archive-ouverte.unige.ch/unige:24793
Disclaimer: layout of this document may differ from the published version.
1 / 1
nature immunology VOLUME 12 NUMBER 8 AUGUST 2011 7 6 1
The migration of neutrophils from the vascular lumen to the extra-
vascular tissue is a vital component of the host’s defense reaction
to injury and infection. To exit the blood circulatory system, neu-
trophils establish a cascade of adhesive interactions with endothelial
cells (ECs) lining the lumen of venular walls and ultimately breach
the endothelium in a polarized manner
1,2
. This process involves dis-
tinct cellular responses that begin with the capture of free-flowing
leukocytes from the circulation and the formation of weak adhesive
interactions with ECs, which results in the rolling of leukocytes along
the venular wall. The activation of rolling neutrophils by surface-
bound stimulating factors such as chemokines promotes their firm
attachment to venular walls and intravascular crawling to sites where
the endothelium is eventually breached. These responses are mediated
by a complex series of overlapping molecular interactions involving
selectins and integrins and their respective ligands
1
.
The migration of neutrophils through ECs (transendothelial migra-
tion (TEM)) can occur via junctions between adjacent ECs (paracellu-
lar route)
3,4
, a response that is supported by the active involvement of
numerous EC junctional molecules, such as PECAM-1, CD99, ICAM-2,
ESAM and members of the junctional adhesion molecule (JAM)
family
1,2,5
. In addition, neutrophils can migrate through the body of
endothelial cells (transcellular route)
6
. Electron microscopy studies of
transcellular TEM have triggered many subsequent investigations into
this phenomenon that have used mainly in vitro models; these have
collectively provided insight into the characteristics and mechanisms
of this mode of TEM
7–12
. For example, invasive leukocyte protrusions
seeking permissive sites and EC structures such as caveolae and the
membranous compartment that connects to the cell surface at cell
borders (the lateral border recycling compartment), which acts as
a source of unligated PECAM-1, CD99 and JAM-A, have all been
associated with mechanisms of transcellular leukocyte TEM
2,7,9,13
.
Despite such studies, fundamental aspects of this response, includ-
ing profile, frequency, dynamics and stimulus specificity in direct
comparison with paracellular TEM, have not been investigated in
real time in vivo.
To further examine the mechanisms by which neutrophils breach
venular walls, we have established a confocal intravital microscopy
imaging platform for the three-dimensional observation of leukocyte
transmigration in real time (four-dimensional imaging). The excep-
tional spatial and temporal resolution of the technique and its appli-
cation to the study of neutrophil TEM in inflamed microvessels has
enabled rigorous analysis of key characteristics of both paracellular
and transcellular TEM. Furthermore, because of the enhanced clar-
ity with which leukocyte TEM can be tracked by this method, we
observed neutrophils that unexpectedly migrated through EC junc-
tions in an abluminal-to-luminal direction (for example, showed
‘reverse TEM’ (rTEM)). We observed such disrupted polarized neu-
trophil paracellular TEM responses relatively selectively under con-
ditions of ischemia-reperfusion (I-R), an inflammatory insult that
caused lower expression of JAM-C at EC junctions. The frequency
of these responses was enhanced after pharmacological blockade or
genetic deletion of EC JAM-C. The pathophysiological relevance of
1
William Harvey Research Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, Charterhouse Square, London, UK.
2
Centre Médical Universitaire, Geneva, Switzerland.
3
Centre for Cardiovascular Research, School of Clinical and Experimental Medicine, College of Medicine and
Dentistry, University of Birmingham, Birmingham, UK.
4
Dresden University of Technology, Dresden, Germany.
5
University of Pennsylvania, Philadelphia, Pennsylvania,
USA.
6
These authors contributed equally to this work. Correspondence should be addressed to S.N. (s.nourshargh@qmul.ac.uk).
Received 10 March; accepted 26 May; published online 26 June 2011; doi:10.1038/ni.2062
The junctional adhesion molecule JAM-C regulates
polarized transendothelial migration of neutrophils
in vivo
Abigail Woodfin
1
, Mathieu-Benoit Voisin
1
, Martina Beyrau
1
, Bartomeu Colom
1
, Dorothée Caille
2
,
Frantzeska-Maria Diapouli
3
, Gerard B Nash
3
, Triantafyllos Chavakis
4,6
, Steven M Albelda
5,6
, G Ed Rainger
3,6
,
Paolo Meda
2,6
, Beat A Imhof
2
& Sussan Nourshargh
1
The migration of neutrophils into inflamed tissues is a fundamental component of innate immunity. A decisive step in this
process is the polarized migration of blood neutrophils through endothelial cells (ECs) lining the venular lumen (transendothelial
migration (TEM)) in a luminal-to-abluminal direction. By real-time confocal imaging, we found that neutrophils had disrupted
polarized TEM (‘hesitant’ and ‘reverse’) in vivo. We noted these events in inflammation after ischemia-reperfusion injury,
characterized by lower expression of junctional adhesion molecule C (JAM-C) at EC junctions, and they were enhanced by
blockade or genetic deletion of JAM-C in ECs. Our results identify JAM-C as a key regulator of polarized neutrophil TEM in vivo
and suggest that reverse TEM of neutrophils can contribute to the dissemination of systemic inflammation.
A R T I C L E S
© 2011 Nature America, Inc. All rights reserved.© 2011 Nature America, Inc. All rights reserved.
7 6 2 VOLUME 12 NUMBER 8 AUGUST 2011 nature immunology
A R T I C L E S
our findings was demonstrated by evidence suggesting that rTEM of
neutrophils stemming from a primary site of trauma (for example,
after I-R injury) was associated with the development of injury in a
second organ. Collectively our findings provide direct evidence that
EC JAM-C has a key role in supporting the luminal-to-abluminal
migration of neutrophils in vivo and suggest that rTEM of neutrophils
can contribute to the dissemination of systemic inflammation.
RESULTS
Four-dimensional imaging for analysis of leukocyte TEM in vivo
To accurately investigate the profile and dynamics of leukocyte TEM
in vivo
, we established a four-dimensional imaging platform with
advanced spatial and temporal resolution. A key component of the
successful application of this imaging method was the reproduc-
ible and adequate labeling of EC junctions for in vivo fluorescence
microscopy imaging. As preliminary studies indicated that intrave-
nous injection of fluorescence-labeled monoclonal antibody (mAb) to
PECAM-1 did not result in sufficiently uniform or strong labeling of
EC contacts for accurate tracking of the route of leukocyte transmigra-
tion, it was necessary to develop an alternative protocol. Intrascrotal
administration of directly labeled Alexa Fluor 555–conjugated mAb
390 to PECAM-1, a mAb that does not inhibit leukocyte transmi-
gration
14
, resulted in strong and reliable labeling of EC borders in
cremasteric venules (Fig. 1a and Supplementary Fig. 1). As well as
showing junctional staining, labeled ECs also showed faint and dif-
fuse cell-body expression of PECAM-1 on the luminal and abluminal
surfaces, which did not seem to be cytoplasmic, as indicated by its
lack of exclusion from nuclear regions when ECs were viewed en face
(Supplementary Fig. 1a–c
). Analysis of PECAM-1-deficient tissues
confirmed that both the junctional and nonjunctional endothelial
labeling was specific (Supplementary Fig. 1d).
In addition to labeling ECs, we also labeled some extravascular cells
with this approach, although these cells did not express PECAM-1, as
demonstrated by analysis of tissues from PECAM-1-deficient mice
and use of isotype-matched control mAbs (Supplementary Fig. 1d).
An additional advantage of labeling EC junctions by intrascrotal injec-
tion of labeled mAb to PECAM (rather than the intravenous route) is
that this approach failed to label circulating leukocytes and thus pro-
vided a cleaner and hence more clear-cut mode of tracking the inter-
actions of leukocytes with EC junctions (
Supplementary Fig. 1e,f).
Although we selected mAb 390 as the mAb to PECAM of choice
because it has been shown not to suppress leukocyte TEM in vivo
14
,
we did rigorous analysis to ensure that this new vascular labeling
protocol had no effect on leukocyte TEM in our model. We confirmed
that intravenous or local administration of mAb 390 did not elicit an
inflammatory response by itself and/or did not effect leukocyte trans-
migration as induced by interleukin 1β
(IL-1β). As expected, the func-
tional blocking mAb MEC 13.3 to PECAM (used as whole molecule)
blocked TEM induced by IL-1β (Fig. 1b,c). Intrascrotal injection of
Alexa Fluor 555–labeled F(ab′)
2
fragment of mAb 390 also effectively
labeled EC junctions, but because the dynamics of leukocyte TEM as
assessed by this approach were not substantially different from those
noted when we used the whole mAb (data not shown), we used the
latter in all subsequent experiments. Finally, fluorescence labeling of
EC junctions by the approach described above, in conjunction with
laser excitation of tissues at our standard confocal microscopy set-
tings, did not elicit leukocyte TEM in unstimulated tissues, in contrast
to the TEM observed in IL-1β-treated tissues (data not shown). We
applied the imaging approach and the EC junction–labeling method
described above to lys-EGFP-ki mice, which express enhanced green
fluorescent protein (GFP) driven by the promoter of the gene encod-
ing lysozyme M and thus have endogenously labeled neutrophils and
monocytes
15
. This allowed detailed spatiotemporal analysis of leuko-
cyte TEM (Fig. 1a and Supplementary Videos 1 and 2).
Inflammation triggers mainly paracellular TEM in vivo
To mimic and analyze the effects of physiological and pathological
insults on leukocyte TEM, we subjected cremaster muscles to the
following three different types of stimuli: the proinflammatory
cytokine IL-1β; the chemotactic formylated tripeptide fMLP; or
I-R injury. Within the in vivo test periods (~2–4 h), the reactions
induced by these stimuli were neutrophilic in nature, as indicated by
published electron microscopy studies
16
and analysis of infiltrates in
stimulated tissues by immunofluorescence staining (data not shown).
Figure 1 Development of a four-dimensional
imaging platform for the analysis of leukocyte
TEM in vivo. (a) Confocal intravital microscopy
of cremasteric venules of lys-EGFP-ki mice
(green leukocytes) immunostained in vivo for EC
junctions by intrascrotal injection of Alexa Fluor
555–labeled mAb 390 to PECAM-1 (red) and
stimulated for 120 min by intrascrotal injection
of IL-1β, followed by surgical exteriorization
and capture of images in vivo at intervals of
1 min for a period of 90 min (from 120 to
210 min after IL-1β injection), showing the
development of an inflammatory response in a
post-capillary venular segment (Supplementary
Video 1). Original magnification, ×40. Scale bar,
10 µm. (b,c
) Brightfield intravital microscopy
of leukocyte adhesion and transmigration
in wild-type mice given no pretreatment
(No mAb) or pretreated intravenously (b) or
intrascrotally (c) with mAb 390, mAb MEC
13.3 or IgG2b isotype-matched control
mAb, then left untreated (saline) or given
intrascrotal administration of IL-1β, followed
by exteriorization of tissues 4 h later. *P < 0.05
and **P < 0.01, IL-1β versus saline and ***P < 0.001 (analysis of variance (ANOVA)). Data are representative of six experiments (a) or are from three to
eight (b) or three to five (c) experiments per group (one mouse per experiment; error bars (b,c), s.e.m.).
60
120 min
160 min
210 min
**
**
**
**
*
*
*
*
*
***
*
50
40
30
Adherent cells (500 µm)
20
10
0
35
30
25
20
15
Transmigrated cells (500 µm)
10
5
0
Saline
No mAb
IL-1β
Saline
IL-1β
Adherent cells (500 µm)
0
25
50
75
Transmigrated cells (500 µm)
0
10
20
30
Saline SalineIL-1β IL-1β
Control mAb (i.v.)
mAb 390 (i.v.)
mAb MEC 13.3 (i.v.)
No mAb mAb 390 (i.s.)
ba
c
© 2011 Nature America, Inc. All rights reserved.© 2011 Nature America, Inc. All rights reserved.
nature immunology VOLUME 12 NUMBER 8 AUGUST 2011 7 6 3
A R T I C L E S
In straight venular segments, the predomi-
nant mode of TEM was paracellular, where
we noted the formation of a pore at EC con-
tacts between two or multiple adjacent cells
(Fig. 2a and Supplementary Videos 2 and 3).
A minority of TEM events analyzed (~10 %) occurred via nonjunc-
tional routes; during these, we noted leukocytes breaching the body
of the endothelium, which resulted in the transient formation of pores
in the cell body (Fig. 2b and Supplementary Video 4). The location
and size of both the paracellular and transcellular pores were diverse
(additional examples of such TEM events with linear intensity profiles
for each pore type, Fig. 2c–f). The linear intensity profiles showed a
paracellular pore immediately flanked by junctions with abundant
PECAM-1 labeling (Fig. 2e), whereas a transcellular pore formed in
the EC body with PECAM-1 labeling of lower intensity (Fig. 2f). We
analyzed the frequency and dynamics of neutrophil TEM via junc-
tional and nonjunctional routes for all three stimuli and found that
~90% of the observed TEM events were via the paracellular route,
with no significant difference between bicellular or multicellular EC
junctions or between different stimuli (Fig. 2g). The mean duration of
each type of TEM response (bicellular, multicellular or transcellular)
was ~6 min, with no significant difference between routes or stimuli
(Fig. 2h). Collectively these results show that in our model, TEM
induced by a variety of distinct proinflammatory stimuli occured pre-
dominantly via the paracellular route, with no significant difference
in the profile or dynamics of the observed responses.
Inflammation triggers several forms of paracellular TEM
Although paracellular was the predominant mode of leukocyte
TEM, we observed various forms of this response. Most paracellular
TEM events involved the normal passage of leukocytes through EC
junctions in a luminal-to-abluminal direction with no pause, but
a smaller proportion of transmigrating leukocytes showed either
rTEM or what seemed to be a multidirectional or ‘hesitant’ mode
of paracellular TEM (Fig. 3). In ‘hesitant’ TEM, leukocytes moved
back and forth in the EC junction several times (about two to three
oscillations) before finally completing migration into the sub-EC
space (Fig. 3a and Supplementary Video 5; additional examples of
hesitant TEM, Supplementary Fig. 2 and Supplementary Videos 6
and 7). In rTEM, leukocytes migrated through EC junctions in the
abluminal-to-luminal direction, disengaged from the junction and
crawled away from the junction across the luminal surface of the
endothelium (Fig. 3b and Supplementary Video 8). We analyzed
the frequency of these various forms of paracellular TEM in tissues
stimulated with IL-1β, fMLP and I-R injury and noted different pro-
files with different stimuli (Fig. 3c
). IL-1β-stimulated tissues showed
relatively little rTEM or hesitant TEM (~3% of all paracellular events
quantified), whereas in tissues subjected to I-R injury, ~15% of para-
cellular TEM events were hesitant TEM or rTEM. The difference in
the frequency of these responses under different inflammatory con-
ditions was not governed by the overall magnitude of the inflamma-
tory responses, as there was no substantial difference in the absolute
number of leukocytes that transmigrated in response to IL-1β or
I-R injury (Supplementary Fig. 3). Quantification of the duration
of the various forms of paracellular TEM in tissues stimulated by I-R
Paracellular TEM Transcellular TEM
0 min 3 min
During paracellular TEM After paracellular TEM After transcellular TEMDuring transcellular TEM
250
5 min 7 min
0 min 4 min 7 min 8 min
*
*
***
*
**
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
a b
c
e
g h
f
d
200
150
100
Fluorescence intensity
(gray values)
Fluorescence intensity
(gray values)
Fluorescence intensity
(gray values)
50
60
0
0 5
†
†
10
JN
JN
Pore
15 20
Distance (µm)
250
200
150
100
50
0
0 5 10
JN
15 20
Distance (µm)
Bicellular junction
Multicellular junction
Nonjunctional
250
200
150
100
50
0
0 5 10
JN
JN
Pore
15 20
Distance (µm)
Fluorescence intensity
(gray values)
250
200
150
100
50
0
0 5 10
JN
JN
15 20
Distance (µm)
50
40
TEM events (%)
30
20
10
0
IL-1β
fMLP I-R
10
8
6
TEM duration (min)
4
2
0
IL-1β
fMLP I-R
Figure 2 Neutrophil paracellular and
transcellular TEM in vivo. (a) Paracellular TEM
of a leukocyte (*; top row) and its associated
transient junctional pore formation (bottom row)
in IL-1β-stimulated, PECAM-1-labeled tissues
(red) of lys-EGFP-ki mice (leukocytes, green;
time (below images) relative to Supplementary
Video 3). (b) Transcellular TEM through ECs with
no disruption of PECAM-1-enriched junctions,
in tissues as in a (top row); below, false-color
images of the PECAM-1 channel (white, high
intensity; blue, low intensity) for visualization
of the transcellular pore (arrows); time (below
images) relative to Supplementary Video 4.
(c,d) Paracellular TEM (c) and transcellular
TEM (d) in tissues as in a. Right, transcellular
pores in close proximity to EC junctions without
disruption of PECAM-1-labeled junctions.
Dotted yellow lines indicate areas analyzed
further in e,f. Scale bars (a–d), 10 µm.
(e,f) Linear intensity profiles of the PECAM-1
channel (EC; red) and GFP (leukocyte; green)
of TEM events along the dotted lines in c,d;
intensity profiles after TEM illustrate pore
closure. JN, junction. (g,h
) Frequency
(g) and duration (h) of TEM events induced
by IL-1β, fMLP or I-R. *P < 0.01 and
**P < 0.001, nonjunctional versus bicellular
and ***P < 0.05 and
†
P < 0.001, nonjunctional
versus multicellular (ANOVA). Data are
representative of four to seven experiments (a)
or are from four to seven experiments with >103
TEM events per group (e–h; one mouse per
experiment; error bars (g,h), s.e.m.).
© 2011 Nature America, Inc. All rights reserved.© 2011 Nature America, Inc. All rights reserved.
7 6 4 VOLUME 12 NUMBER 8 AUGUST 2011 nature immunology
A R T I C L E S
indicated that hesitant TEM and rTEM were significantly slower
than normal paracellular responses (Fig. 3d).
Neutrophils show disrupted polarized TEM
We next sought to further investigate the profile of hesitant TEM and
rTEM, collectively called ‘disrupted TEM’ here, under conditions of
I-R injury. Specifically, as we observed these events in I-R-injured
tissues from lys-EGFP-ki mice, which express GFP in both neutrophils
and monocytes
15
, it was important to elucidate which leukocyte sub-
type had disrupted TEM responses. We investigated the contribution
of neutrophils and monocytes to the observed TEM events by various
approaches, analyzing neutrophil-depleted lys-EGFP-ki mice (Fig. 4a
and Supplementary Fig. 4a) and the differentiation of neutrophils
from monocytes on the basis of significant differences in their GFP
intensities (Fig. 4b,c, Supplementary Fig. 4b and Supplementary
Results) and, finally, through the use of CX3CR1-GFP-ki mice, which
express enhanced GFP in all monocytes but not in neutrophils
17
(
Supplementary Fig. 4c). In these last mice, I-R injury elicited very
little normal paracellular TEM of GFP
+
cells, and we detected no hesi-
tant TEM or rTEM in >20 vessels analyzed. Collectively, these findings
indicated that neutrophils were the sole participants in hesitant TEM
and rTEM in our I-R injury model.
I-R injury lowers JAM-C expression at EC junctions
Having characterized the responses of disrupted TEM in terms of
frequency, stimulus specificity, dynamics and leukocyte subtype,
we next addressed the mechanism associated with these events. We
hypothesized that hesitant TEM and rTEM were associated with
disrupted expression of EC junctional mole-
cules. To address this possibility, we used
immunofluorescence staining and confocal
microscopy to investigate the expression of
vascular endothelial cadherin (VE-cadherin)
and PECAM-1, key EC junctional molecules
linked to maintenance of the integrity of EC junctional contacts
18
and the mediation of leukocyte TEM, respectively
1,2,19
, in control,
IL-1β-stimulated and I-R-injured tissues. Furthermore, as JAM-C
supports the polarized migration of monocytes through cultured ECs
in vitro
20
, we also analyzed the expression of this adhesion molecule.
Under conditions of I-R injury, but not in response to IL-1β, cell sur-
face expression of JAM-C was lower at EC junctions. We noted no
change in VE-cadherin or PECAM-1 with I-R injury or in response to
IL-1β (Fig. 5a,b). The loss of JAM-C at EC junctions after I-R injury
was partially prevented by pretreatment of mice with superoxide dis-
mutase and catalase (Fig. 5b), which indicated that reactive oxygen
intermediates contributed to the lower junctional expression of JAM-C
in response to I-R injury.
We also investigated the expression of JAM-C in inflamed tissues by
immunoelectron microscopy (Fig. 5c and Supplementary Fig. 5). This
approach confirmed published findings that JAM-C can be detected on
ECs at the following three locations: junctional membranes, nonjunc-
tional membranes and cytosolic vesicles
21
. In control (sham-operated
and saline-injected) tissues, JAM-C was distributed almost equally
among those three locations. In tissues subjected to I-R injury, there
was less labeling of JAM-C at junctions and cytosolic vesicles, whereas
labeling of the nonjunctional membrane was greater. In contrast and in
agreement with our immunofluorescence staining studies, we noted no
change in JAM-C expression in IL-1β-stimulated tissues, as analyzed by
immunoelectron microscopy (Fig. 5c). Together these findings demon-
strated the ability of I-R injury, but not of IL-1β, to induce lower expres-
sion of JAM-C at EC junctions, findings that were in line with the almost
undetectable disrupted paracellular TEM induced by IL-1β (Fig. 3c).
Lumen
Lumen
0 min 5 min 13 min 19 min 24 min 27 min
0 min
100
90
80
20
Paracellular TEM events
(% of total)
15
10
5
0
Paracellular TEM duration
(min)
0
10
20
30
IL-1β
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
**
*
fMLP I-R
Normal
TEM
Hesitant
TEM
rTEM
4 min 6 min
Normal TEM
Hesitant TEM
rTEM
9 min 13 min 22 min
c
b
a
d
Figure 3 Disrupted forms of polarized
paracellular TEM. (a) Time-lapse images of
a GFP-expressing leukocyte (*) undergoing
hesitant paracellular TEM: top, transverse
section of venule; middle, luminal view; bottom,
sub-EC segments of the migrating leukocyte in
light green with dashed outline (time (below
images) relative to Supplementary Video 5;
additional examples, Supplementary Fig. 2 and
Supplementary Videos 6 and 7). (b) Time-lapse
images of a leukocyte undergoing rTEM as it
migrates through a bicellular junction in an
abluminal-to-luminal direction, disengages from
the junction and crawls away on the luminal
surface: top, transverse section of venule;
bottom, luminal view (time (below images)
relative to Supplementary Video 8). (c) Frequency
of normal, hesitant and reverse paracellular TEM
events induced by IL-1β, fMLP or I-R, presented
as frequency among total paracellular TEM
events. (d) Duration of normal, hesitant and rTEM
events in tissues injured by I-R. Scale bars (a,b),
10 µm. *P < 0.001, normal versus disrupted
(hesitant TEM and rTEM) and **P < 0.001,
hesitant TEM versus rTEM (ANOVA). Data are
representative of seven experiments (a,b) or are
from four to seven experiments with >103 TEM
events per group (c,d; one mouse per experiment;
error bars (c,d), s.e.m.).
© 2011 Nature America, Inc. All rights reserved.© 2011 Nature America, Inc. All rights reserved.
