This information is current as
of June 17, 2017.
LFA-1 and Mac-1 Define Characteristically
Different Intralumenal Crawling and
Emigration Patterns for Monocytes and
Neutrophils In Situ
Ronen Sumagin, Hen Prizant, Elena Lomakina, Richard E.
Waugh and Ingrid H. Sarelius
Supplementary
Material
References
Subscription
Permissions
Email Alerts
http://www.jimmunol.org/content/suppl/2010/10/29/jimmunol.100163
8.DC1
This article cites 48 articles, 31 of which you can access for free at:
http://www.jimmunol.org/content/185/11/7057.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
J Immunol 2010; 185:7057-7066; Prepublished online 29
October 2010;
doi: 10.4049/jimmunol.1001638
http://www.jimmunol.org/content/185/11/7057
The Journal of Immunology
LFA-1 and Mac-1 Define Characteristically Different
Intralumenal Crawling and Emigration Patterns for
Monocytes and Neutrophils In Situ
Ronen Sumagin,* Hen Prizant,* Elena Lomakina,† Richard E. Waugh,†
and Ingrid H. Sarelius*
R
ecruitment of circulating leukocytes from blood vessels
into the surrounding tissue is characteristic of the inflammatory response and is essential for host defense. The
first line of defense is provided primarily by neutrophils through
phagocytosis and the release of cytotoxic molecules (1). Along
with neutrophils, monocyte emigration has also been documented
in the early stages of the inflammatory response (2, 3). Transendothelial migration (TEM) of both neutrophils and monocytes
requires several sequential steps such as rolling (4), adhesion (5),
and intralumenal crawling (6, 7). The initial rolling of leukocytes
is mediated by multiple members of the selectin family and the
glycosylated molecules expressed on the leukocyte surface (8).
Subsequent adhesion and crawling are primarily mediated via the
interactions between the members of cell adhesion molecule and
b1/b2 integrin families (9); however, the specificity of these interactions differs between neutrophils and monocytes (10, 11).
Neutrophil adhesion to activated endothelium is mainly mediated
*Department of Pharmacology and Physiology and †Department of Biomedical Engineering, University of Rochester, Rochester, NY 14642
Received for publication May 17, 2010. Accepted for publication September 27,
2010.
This work was supported by National Institutes of Health Grants RO1 HL75186 and
PO1 HL18208.
Address correspondence and reprint requests to Dr. Ingrid H. Sarelius, Department
of Pharmacology and Physiology, University of Rochester Medical Center, 601
Elmwood Avenue, Rochester, NY 14642. E-mail address: ingrid_sarelius@urmc.
rochester.edu
The online version of this article contains supplemental material.
Abbreviations used in this paper: EC, endothelial cell; fps, frames per second; KO,
knockout; TEM, transendothelial migration; WT, wild-type.
Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00
www.jimmunol.org/cgi/doi/10.4049/jimmunol.1001638
by aLb2 (LFA-1) (6, 12), but the lumenal crawling that follows
adhesion and spreading is aMb2 (Mac-1) dependent (6). In contrast, recent mouse models showed that monocytes can also use
a4b1 (VLA-4) for firm adhesion (13, 14) and LFA-1 for intralumenal crawling (15). Leukocyte TEM occurs in specific, localized endothelial regions [portals (16)], which are primarily located
at endothelial cell (EC) junctions (16); thus, presumably, intralumenal crawling allows leukocytes to get to these locations. The
nature of these locations is still undefined; however, tricellular
junctional regions that are enriched in ICAM-1 (16), as well as
localities expressing lower levels of key extracellular matrix proteins comprising the basal lamina (17), have been identified as integral to these locations.
There is emerging evidence for neutrophil-mediated monocyte
recruitment (18), implying that the recruitment of these leukocyte
subpopulations might occur in sequence (19), but whether neutrophils and monocytes use similar mechanisms for TEM and
whether this occurs at the same locations is not known. Indeed,
differences in the way neutrophils and monocytes penetrate the
basement membrane were recently documented (2). Similarly, inhibition of either Mac-1 or LFA-1 has been associated with decreased leukocyte TEM (12, 20), but whether this is due to a direct
role of these integrins in guiding leukocytes through the endothelium, or whether this is due to the inability of leukocytes to
properly adhere or to crawl to the specific TEM locations is yet to
be determined.
Intralumenal crawling by leukocytes was first described in
isolated cell systems (21) and subsequently confirmed in situ in
blood-perfused microvessels (7). Intralumenal crawling is essential for TEM, as it enables leukocytes to get to the portal locations
that accommodate leukocyte passage. Preventing leukocytes from
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
To exit blood vessels, most (∼80%) of the lumenally adhered monocytes and neutrophils crawl toward locations that support
transmigration. Using intravital confocal microscopy of anesthetized mouse cremaster muscle, we separately examined the
crawling and emigration patterns of monocytes and neutrophils in blood-perfused unstimulated or TNF-a–activated venules.
Most of the interacting cells in microvessels are neutrophils; however, in unstimulated venules, a greater percentage of the total
monocyte population is adherent compared with neutrophils (58.2 6 6.1% versus 13.6 6 0.9%, adhered/total interacting), and
they crawl for significantly longer distances (147.3 6 13.4 versus 61.8 6 5.4 mm). Intriguingly, after TNF-a activation, monocytes
crawled for significantly shorter distances (67.4 6 9.6 mm), resembling neutrophil crawling. Using function-blocking Abs, we show
that these different crawling patterns were due to CD11a/CD18 (LFA-1)- versus CD11b/CD18 (Mac-1)-mediated crawling.
Blockade of either Mac-1 or LFA-1 revealed that both LFA-1 and Mac-1 contribute to monocyte crawling; however, the LFA1–dependent crawling in unstimulated venules becomes Mac-1 dependent upon inflammation, likely due to increased expression of
Mac-1. Mac-1 alone was responsible for neutrophil crawling in both unstimulated and TNF-a–activated venules. Consistent with
the role of Mac-1 in crawling, Mac-1 block (compared with LFA-1) was also significantly more efficient in blocking TNF-a–
induced extravasation of both monocytes and neutrophils in cremaster tissue and the peritoneal cavity. Thus, mechanisms underlying leukocyte crawling are important in regulating the inflammatory responses by regulating the numbers of leukocytes that
transmigrate. The Journal of Immunology, 2010, 185: 7057–7066.
7058
Materials and Methods
Animals
Experiments were performed on male wild-type (WT) C57BL6J mice (The
Jackson Laboratory, Bar Harbor, ME) between the ages of 12 and 15 wk.
When indicated, inflammation was induced by local treatment with mouse
rTNF-a (intrascrotal injection, 0.5 mg TNF-a in 0.25 ml saline; SigmaAldrich, St. Louis, MO) 3 h prior to the start of the surgical preparation
(25). Observations were made 4–6 h after the TNF-a injection. All mice
were used according to protocols approved by the University of Rochester
Institutional Review Board (Rochester, NY).
Intravital microscopy
Mice were anesthetized with an initial dose of sodium pentobarbital (65 mg/
kg i.p.). Supplemental anesthetic was administered as needed throughout
the experiment via a catheter inserted into the jugular vein. An endotracheal
tube was inserted to insure a patent airway during the experiment, and the
animal was kept warm by placing it on a warmer. The right cremaster
muscle was exteriorized and gently pinned over a quartz pedestal for visualization by microscopy. During preparation and observation, the tissue
was continuously superfused with warmed physiological solution with the
following composition (in mM): NaCl, 131.9; KCl, 4.7; CaCl2, 2.0; MgSO4,
1.2; and NaHCO3, 18 (pH 7.4), at 36˚C, and equilibrated with gas containing 0% O2, 5% CO2, and 95% N2 to maintain tissue PO2 ,15 torr (26).
Observations were made using an Olympus BX61WI microscope with an
Olympus PlanF1 immersion objective (320, 0.65 numerical aperture, or
340, 0.95 numerical aperture; Olympus, Center Valley, PA). Transilluminated images were acquired via a charge-coupled device camera
(Dade MTI CD72, DageMTI, Michigan City, IN). Fluorescence images
were acquired by illuminating the tissue with a 50 mW argon laser and
imaging with a Nipkow disk confocal head (CSU 10, Yokogawa Electric,
Tokyo, Japan) attached to an intensified CCD camera (XR Mega 10,
Stanford Photonics, Palo Alto, CA): laser power and camera gain settings
were unchanged throughout all the experiments. Images were either digitally acquired or recorded to a DVD recorder (SONY DVO100MD, Sony,
Tokyo, Japan) at 30 frames per second (fps) for offline analysis. All protocols were completed within 2 h of the initial observation. Upon completion of the protocols, the animal was euthanized by anesthetic overdose.
Flow cytometric analysis
The surface expression of Mac-1 and LFA-1 on monocytes and neutrophils
in mouse circulation under baseline and TNF-a–activated conditions was
assessed using flow cytometry. To do this, mouse blood was diluted 1:6
with PBS to lyse RBCs and washed twice in 10 mM HEPES-buffered
physiological salt solution containing 0.1% BSA. To identify monocytes
and neutrophils respectively, leukocytes were labeled with Abs against F4/
80 and GR-1 conjugated to Alexa 488 (10 mg/ml; Serotec, Raleigh, NC)
together with CD11a, CD11b (20 mg/ml; eBioscience, San Diego, CA), or
CD49d (20 mg/ml; Southern Biotechnology Associates, Birmingham, AL).
A total of 5000 cells per sample were analyzed by flow cytometry (Guava
EasyCyte Mini, Guava Technologies, Hayward, CA) as previously described (27). To estimate the number of Ab-binding sites per cell (molecules/cell), each fluorescence signal was calibrated using Quantum Simply
Cellular Beads (Flow Cytometry Standards, Fishers, IN). A standard suspension of beads containing five different populations with known numbers
of binding sites was labeled to saturation with the Abs used to label the
cells, and the fluorescence signals were quantified. The fluorescence intensity of the cell samples was converted to numbers of binding sites from
this standard solution using software provided by the manufacturer of the
beads. To correct for nonspecific binding, the number of nonspecific sites
identified using isotype control Abs was subtracted from the total number
of sites detected using each specific Ab.
In situ immunofluorescence labeling
In separate experiments, we used either anti-F4/80 or anti–GR-1 Abs
conjugated to Alexa 488 (CI:A3-1 and RGB-8C5, respectively; 1.5 mg/
mouse, i.v.; Serotec) to visualize monocytes and neutrophils, respectively.
GR-1 Ag is widely used to identify circulating neutrophils (6, 28–30);
however, a subset of monocytes has also been shown to be GR-1+/+ (31,
32). Using these markers, we were able to distinguish differences in
crawling behavior and TEM between monocytes and neutrophils. It is indeed possible that in the behavior analysis of GR-1+/+ cells (neutrophils), we
included by default a number of GR-1+/+ monocytes. However, as monocytes constitute ,5% of circulating leukocytes (2), and GR-1+/+ monocytes
constitute only a subset of the total monocyte population, and as we analyzed .80 cells in each experimental group, this small fraction will not
significantly affect our measurement and conclusions. Adhesion and
crawling of neutrophils stained with RGB-8C5 Ab was also compared with
the behavior of neutrophils stained with anti-neutrophil Ab (7/4, 1.5 mg/
mouse, i.v.; Cedarlane Laboratories, Burlington, NC) (data not shown), as
well as with an mAb 1A8 (anti–Ly-6G, 1.5 mg/mouse, i.v.; Biolegend, San
Diego, CA) (Supplemental Fig. 5), which specifically binds Ly-6G and was
previously used to identify granulocytes (33). For both 7/4 and 1A8 Abs, no
significant differences in neutrophil adhesion and crawling were found
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
crawling significantly decreased TEM (6, 16). Furthermore, leukocytes that were able to migrate across the endothelium migrated
at nonoptimal locations and required significantly longer time (6).
Due to the EC morphology, rolling leukocytes land by default on
EC junctions (7, 16), but the majority immediately engage in
crawling toward other junctional locations (6, 16). In most cases,
crawling occurs in random directions (independent of flow direction), although this random crawling is lost in mice lacking
functional Vav1, which is a major regulator of actin organization
during leukocyte migration (22). Similarly, proinflammatory stimuli such as fMLP (16) induce leukocyte crawling that is parallel or
perpendicular to flow but rarely against the flow direction, indicating a degree of directionality. fMLP or TNF-a activation
causes rearrangement of EC surface adhesion molecules, such as
ICAM-1 (16), and, in turn, blockade of ICAM-1, which is essential
for leukocyte crawling (6), results in a loss of fMLP-induced directionality in the remaining fraction of crawling leukocytes (16).
This suggests first that some locations are better equipped to accommodate leukocyte passage than others and second that the most
likely factors to drive leukocyte crawling are the distribution of
adhesion molecules and as yet unidentified chemotactic signals.
The crawling patterns of the different leukocyte subpopulations
(neutrophils and monocytes) are also different. For example, the
reported crawling distances for monocytes are significantly longer
[up to 250 mm (15)] than those reported for neutrophils [10–40
mm (22)]. Similarly, monocytes can crawl in unstimulated venules
(15) and have the ability to reverse migrate [from tissue into the
blood vessels (23)], both consistent with their suggested patrolling
(healthy tissue monitoring) function (15). To our knowledge, there
is not yet evidence for neutrophils to either crawl in unstimulated
tissue or to exhibit reverse migration. Finally, a significantly greater
number of monocytes undergo transmigration compared with
neutrophils [relative to the number of these leukocyte subtypes in
blood, ∼4 and 14%, respectively (2)]. As crawling is essential for
TEM (24), it is likely that both the expression of the specific
molecules mediating this crawling as well as the crawling patterns
will impact the number of transmigrated leukocytes in a defined
time period.
In the current study, we used intravital fluorescence confocal
microscopy to separately examine the crawling and emigration
patterns of monocytes and neutrophils in blood-perfused unstimulated or TNF-a–activated venules. Our goal was to determine whether observed differences in monocyte and neutrophil
crawling behavior result from mediation by different b2 integrins and
whether this consequently affects TEM. We show that in unstimulated venules, monocytes but not neutrophils are actively engaged
(adhered/crawling) with the vessel wall. Although neutrophils
predominately use Mac-1 for crawling, monocytes interchangeably
use both LFA-1 (in unstimulated venules) and Mac-1 (in inflammation) for crawling. Importantly, we determined that the
distance traveled for LFA-1–mediated crawling is characteristically
longer than that mediated by Mac-1, which is consistent with the
concept that this crawling behavior is used to survey the healthy
tissue. This indicates that mechanisms underlying leukocyte
crawling are important in regulating the inflammatory responses by
regulating the numbers of transmigrated leukocytes.
LEUKOCYTE CRAWLING AND EMIGRATION IN SITU
The Journal of Immunology
7059
compared with the data from GR-1 labeling (Figs. 3, 4). Likewise, to
confirm that labeling with F4/80 Ab detects all circulating monocytes, we
compared the results obtained from these experiments to a mouse strain,
CX3CR1eGFP/+, that exhibits endogenously enhanced GFP-labeled monocytes. We found no significant differences in the interacting monocyte
fraction (Supplemental Fig. 4B), adhesion, and TEM (Supplemental Fig.
4A) compared with the numbers obtained with F4/80 staining (Figs. 3, 4).
This confirms that under the conditions of our studies, F4/80 epitope is
present on all blood monocytes and can be used for their identification. As
shown in Supplemental Fig. 1, GR-1 and F4/80 Abs at the concentrations
used in this work did not affect leukocyte rolling, adhesion, and TEM. We
also determined that the selected concentration of 1.5 mg/mouse is optimal
for both GR-1 and F4/80 labeling (Supplemental Fig. 2) and that within the
2 h time of experimental procedures, GR-1 Ab had no depleting effect
(Supplemental Fig. 3). To quantify rolling, adhesion, and crawling, all Abs
were given i.v. in 50 ml saline 5 min prior to observations via a second
catheter inserted into the jugular vein. To quantify TEM, all Abs were given
via tail vein in 200 ml saline. For all blocking experiments, LFA-1– and
Mac-1–blocking Abs (M17/4, M1/70, respectively; 50 mg/mouse; eBioscience) as well as VLA-4–blocking Ab (PS/2, Southern Biotechnology
Associates) were given together with the labeling Abs either via the jugular
vein or by tail vein injection.
To quantify leukocyte transmigration into the peritoneal cavity, a separate
group of mice were given an i.p. injection of TNF-a (100 ng in 50 ml
saline). Four hours later, peritoneal cavities of anesthetized mice were
lavaged with 3 ml PBS, and white cells were recovered and counted using
a hemocytometer. Leukocyte differentials were determined on 100 ml
cytospins stained with Diff-Quik (Dade Behring, Newark, DE). For all
blocking experiments, the appropriate Ab was injected in 200 ml saline via
the tail vein 10 min prior to i.p. injection of either saline or TNF-a and 4 h
prior to the lavage protocol.
Analysis
Leukocyte–EC interactions. Rolling leukocytes were defined as any leukocytes observed translating along the vessel wall in continuous contact
with the endothelium (34). Delivered leukocytes were defined as all leukocytes seen in the vicinity of the wall independently of whether they were
rolling or carried in the free stream. The number of rolling leukocytes on
the vessel wall was calculated by counting leukocytes rolling past a line
perpendicular to the vessel axis per 40-s time interval. All leukocytes that
remained stationary or did not exceed a displacement of .8 mm (one
leukocyte diameter) during 30 s were considered adhered. Leukocytes that
exceeded a displacement of .8 mm were considered crawling. For all
treatment groups, only leukocytes that were attached to the endothelium
and were able to crawl were analyzed. To quantify leukocyte crawling,
each venular site was observed and recorded for 40 min at 30 fps. The
original movies were time lapsed to 0.33 fps for offline analysis. Crawling
distances were obtained by measuring the path length of each crawling
leukocyte during the 40 min observation time. The confinement ratio was
defined as displacement/total path length (1 = straight line). In experiments
in which LFA-1 and Mac-1 blocking Abs were used, an appropriate control
isotype or a combination of isotypes was tested. Both control isotypes had
no significant effect on leukocyte behavior and were not different from
each other (data not shown); hence, for simplicity in Figs. 3, 4, and 7, all
data are presented against IgG2b (a control isotype for Mac-1). For TEM,
leukocytes were counted in the extravascular tissue within 50 mm of the
vessel/100-mm length vessel segments. All TEM counts were normalized
to 10,000 mm2.
Statistical significance was assessed by Student t test or by one way
ANOVA with Newman-Keuls multiple comparison test using GraphPad
Prism (version 4.0, GraphPad, San Diego, CA). Statistical significance was
set at p , 0.05.
Results
The majority of monocytes but not neutrophils are adhered and
crawling in uninflamed venules
Leukocytes roll abundantly in unstimulated, surgically prepared
venules in situ, but rarely adhere (35). As rolling is an essential step
in the leukocyte recruitment cascade, we asked whether this step is
equally important for both neutrophils and monocytes. Using anti–
GR-1 and F4/80 Abs to identify neutrophils and monocytes, respectively, we quantified their interactions with ECs. GR-1 and F4/
80 staining indicated that in control venules, ∼60% of the total
leukocyte population interacting with the vessel wall (rolling, adhered, and crawling) are neutrophils, and ,20% are monocytes
(the rest presumably are lymphocytes). Both neutrophils and monocytes were observed rolling, as summarized in Table I. Both monocytes and neutrophils maintain a round shape while rolling, but
following adhesion undergo flattening to begin to crawl. The round
shape of a representative rolling neutrophil and a flattening neutrophil that is transitioning from adhesion to crawling are illustrated in Fig. 1D, as captured by transmission electron microscopy
(upper panels) and in situ bright field microscopy (bottom panels).
Interestingly, whereas only a small fraction of interacting neutrophils (13.6 6 0.9%; Fig. 1A) became firmly adhered (Fig. 1C,
indicated by the arrows) and subsequently exhibited crawling in
unstimulated venules, the majority (58.2 6 6.1%; Fig. 1A) of
interacting monocytes were found adhered (Fig. 1C, indicated by
the arrows), and 47.7 6 5.0% of these adhered cells exhibited
crawling (Fig. 1B). These differences in the number of interacting
monocytes and neutrophils are illustrated by the representative
snapshots of the fields of view of selected venules in which monocytes were stained for F4/80 (Fig. 1C, upper panel) and neutrophils
for GR-1 (Fig. 1C, bottom panel). These findings clearly indicate
that in unstimulated venules, monocytes but not neutrophils are
actively engaged with the vessel wall, confirming a previous report
in which monocytes were suggested to exhibit patrolling behavior
(random crawling) in control venules (15).
Monocyte crawling patterns become neutrophil-like upon
TNF-a activation
TNF-a significantly increases leukocyte adhesion in situ (35, 36).
We show in this study that this is true for both monocytes and
neutrophils. The number of adhered neutrophils following TNF-a
treatment increased ∼3-fold (from 4.8 6 0.8 to 14.4 6 1.4 cells/
field) and monocytes ∼2-fold (from 2.1 6 0.4 to 4.3 6 0.6 cells/
field; Fig. 2A). Thus, as in unstimulated venules, in TNF-a–acti-
Table I. Leukocyte rolling in unstimulated and TNF-a–activated venules
Condition
Unstimulated
TNF-a
Leukocyte
Subtype
+/+
Neutrophils (GR-1 )
Monocytes (F4/80+/+)
Neutrophils (GR-1+/+)
Monocytes (F4/80+/+)
Delivered
(Cells/40 s)a
Rolling
(Cells/40 s)a
Rolling Fraction
(Percent Total)
Rolling Velocity
(mm/s)
N
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
100
100
100
100
18.2
6.1
6.6
2.4
1.4
0.8
1.1
1.1
13.7
3.9
5.9
2.2
1.7
0.6
1.2
0.7
73.8
65.1
92.7
93.1
2.4
4.5
5.4
5.5
37.9
48.5
4.1
4.3
a
1.5
1.9a
0.3b
1.4b
Values are means 6 SE. Neutrophils and monocytes were fluorescently labeled with GR-1 and F4/80 Abs, respectively, in unstimulated or TNF-a–
activated venules. Delivered cells included all rolling cells and the cells that were not in direct contact with the endothelium but were translating in the
free stream near the wall. Rolling fraction was defined as number of rolling cells/total delivered.
a
Significantly different from each other (p , 0.01).
b
Significantly different from appropriate unstimulated condition (p , 0.01).
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
Leukocyte emigration into the peritoneal cavity
Statistics
7060
LEUKOCYTE CRAWLING AND EMIGRATION IN SITU
nificantly increased (0.83 6 0.03), indicating a more direct path.
The crawling trajectories of 10 representative monocytes (Fig. 2F)
illustrate the typical difference in length and confinement ratio
under control (upper panel) and TNF-a–activated conditions (lower
panel).
Interchangeable roles for LFA-1 and Mac-1 in monocyte
crawling
vated vessels, the majority of the cells recruited to the wall are
neutrophils. As expected from recent work (15, 16), adhered neutrophils and monocytes were observed to crawl intralumenally in
both control and TNF-a–activated venules (Fig. 2B), however,
with characteristically different crawling patterns. In TNF-a–
activated venules, neutrophils exhibited direct crawling (close to
a straight line) with a confinement ratio of 0.83 6 0.03 (confinement ratio is defined as displacement/total path length, 1 =
straight line) and average crawling distance of 61.8 6 5.4 mm
(Fig. 2D, 2E). The small number of neutrophils that were observed
crawling in unstimulated venules exhibited similar crawling patterns. In contrast to neutrophils, monocyte crawling under control
conditions was less direct (confinement ratio of 0.64 6 0.03) and
for significantly longer distances, with an average distance of at
least 147.3 6 13.4 mm (Fig. 2D). This supports the idea (15) that
an important function of monocytes in uninflamed tissue is to
provide constant monitoring. The crawling velocities of both
monocytes and neutrophils were on the order of 10 mm/min and
were not significantly different in unstimulated compared with
TNF-a–activated venules (Fig. 2C). Fig. 2G illustrates the paths
of three crawling monocytes (upper panel) versus three crawling
neutrophils (lower panel) in a representative unstimulated venule
over a 40-min time period. Intriguingly, after TNF-a treatment,
monocyte crawling behavior (the crawling distance and the confinement ratio) changed to closely resemble that of neutrophils.
The average distance that monocytes crawled became significantly
shorter (67.4 6 9.6 mm; Fig. 2D) and the confinement ratio sig-
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 1. The majority of monocytes but not neutrophils adhere and
crawl in uninflamed venules. The interactions of monocytes (stained for
F4/80; 1.5 mg/mouse, i.v.) and neutrophils (stained for GR-1; 1.5 mg/
mouse, i.v.) with the vessel wall were quantified in unstimulated venules.
A, Interacting cells were defined as all cells in immediate contact with the
vessel wall. Field of view is ∼300 mm length. B, The fraction of leukocytes
that exhibited crawling out of total number of interacting cells. For A and
B, n = 11 venules, four mice. C, Representative images of the fields of view
of selected venules where either monocytes (upper panel) or neutrophils
(bottom panel) were immunofluorescently labeled. White arrows indicate
the adhered leukocytes (confirmed from analysis of the movie from which
the frame was extracted). Scale bar, 15 mm. Whereas the majority of
neutrophils exhibit rolling behavior in unstimulated venules, the majority
of monocytes were found adhered or crawling under these conditions. D,
The round shape of a representative rolling neutrophil and a flattening
neutrophil that is transitioning from adhesion to crawling, as captured by
electron microscopy (upper panels, scale bar, 1.8 mm) and in situ bright
field microscopy (bottom panels, scale bar, 10 mm).
Recent studies indicate that monocyte crawling is primarily mediated by LFA-1, whereas neutrophil crawling has been identified
as a Mac-1–mediated phenomenon (6, 15). We hypothesized that
the differences in monocyte crawling patterns in unstimulated
versus TNF-a–activated venules could be attributed to Mac-1–
versus LFA-1–mediated crawling. To test this hypothesis, we
quantified the effect of blockade of LFA-1, Mac-1, or both on
leukocyte crawling in control and TNF-a–activated venules. We
found, as expected (6), that neutrophil crawling is mainly mediated
by Mac-1 in both control and TNF-a–activated venules. However,
in contrast, monocyte crawling is mediated by LFA-1 in unstimulated venules but becomes Mac-1 dependent after TNF-a activation. Blockade of Mac-1 had no significant effect on neutrophil
adhesion in both control and activated venules (Fig. 3A, 3B, respectively); however, the crawling fraction (crawling/adhered) significantly decreased under both conditions (76.2 6 3.9 IgG versus
22.4 6 2.3% Mac-1 block in unstimulated venules and 76.4 6 5.0
IgG versus 32.7 6 5.8% Mac-1 block in TNF-a–activated venules;
Fig. 3C, 3D). In contrast, LFA-1 block significantly attenuated
neutrophil adhesion (4.3 6 0.5 IgG versus 1.9 6 0.5 cells/field
LFA-1 block in unstimulated venules and 13.7 6 1.1 IgG versus
3.6 6 0.7 cells/field LFA-1 block in TNF-a–activated venules; Fig.
3), but did not prevent the remaining fraction from crawling (Fig.
3C, 3D). A combination of anti–Mac-1 and –LFA-1 (Fig. 3C, 3D)
or, alternatively, blockade of CD18 (data not shown) had no additive effect over that achieved by Mac-1 block alone on either
neutrophil adhesion or crawling. Blockade of LFA-1 significantly
reduced monocyte adhesion in unstimulated venules (2.0 6 0.4 IgG
versus 1.0 6 0.3 cells/field LFA-1 block; Fig. 3A), but neither LFA1 nor Mac-1 block affected monocyte adhesion in TNF-a–activated
venules (Fig. 3B). Consistent with earlier work (15), blockade of
LFA-1 but not Mac-1 significantly attenuated monocyte crawling
fraction (by ∼50%; Fig. 3C), but this only occurred in control
venules. When the tissue was activated with TNF-a, blockade of
LFA-1 failed to significantly affect monocyte crawling fraction
(Fig. 3D), whereas blockade of Mac-1 resulted in a .40% decrease
in monocyte crawling (Fig. 3D). We further confirmed this using
Mac-1 knockout (KO) mice. Consistent with the results obtained
with the Mac-1 blocking Ab, monocyte crawling in unstimulated
venules in Mac-1 KO mice was unaffected by the lack of functional
Mac-1; however, only ∼36% of all adhered monocytes were able to
crawl in TNF-a–activated vessels (Supplemental Fig. 6).
Interestingly, blockade of both Mac-1 and LFA-1 as well as CD18,
(data not shown) under either unstimulated (Fig. 3C) or TNF-a–
activated (Fig. 3D) conditions produced a further decrease in
monocyte crawling fraction, suggesting that in contrast to neutrophils, monocytes can use either of the integrins under specific
circumstances. As the combination of LFA-1 and Mac-1 block in
inflamed venules did not significantly decrease monocyte adhesion
and only partially blocked the ability of monocytes to crawl, we
asked whether VLA-4 contributes to monocyte–EC interactions. In
control venules, VLA-4 block had no effect on monocyte adhesion
and crawling (Fig. 3A, 3C). This is consistent with our previous
observation (35) in which we found that VCAM-1 (endothelial
ligand for VLA-4) is almost undetectable on resting endothelium.
This also argues that LFA-1 indeed mediates leukocyte crawling
The Journal of Immunology
7061
under these conditions without a contribution from VLA-4. Intriguingly, in TNF-a–activated vessels, VLA-4 block significantly
reduced monocyte adhesion; however, its effect on the monocyte
crawling fraction (Fig. 3C, 3D), the crawling distance, and the
crawling velocity (Fig. 4) was not significant. Blockade of VLA-4
had no significant effect on neutrophil adhesion and crawling in
both control and TNF-a–activated vessels. Together, these findings
suggest that both LFA-1 and Mac-1, but not VLA-4, contribute to
monocyte crawling; however, the dominant role in mediating
monocyte crawling changes from LFA-1 in unstimulated venules to
Mac-1 upon inflammation.
LFA-1–mediated crawling distance is characteristically longer
than that of Mac-1
We showed that the characteristically long (147.3 6 13.4 mm; Fig.
2D) monocyte crawling distance in unstimulated venules was LFA-1
dependent (Fig. 3C). In contrast, the significantly shorter distance
crawled by monocytes in TNF-a–activated venules (67.4 6 9.6 mm;
Fig. 2D) was mainly Mac-1 dependent (Fig. 3D). Moreover, monocyte crawling distances in TNF-a–activated venules were similar to
distances crawled by neutrophils in either unstimulated or TNF-a–
activated vessels, in which crawling is also mediated by Mac-1. This
indicates that the LFA-1–mediated crawling distance is characteristically longer than that of Mac-1. To further test this, we measured
the crawling distances of monocytes and neutrophils in the presence
of either LFA-1– or Mac-1–blocking Abs. In unstimulated venules
following LFA-1 block, the small fraction of adhered monocytes that
were able to crawl (Fig. 3A) crawled for significantly shorter distances compared with those in untreated venules or venules treated
with control isotype Abs (61.0 6 8.8 LFA-1 block versus 144.0 6
10.2 mm IgG; Fig. 4A) and with significantly slower crawling velocities (Fig. 4C). We conclude that this crawling was likely mediated
by Mac-1 because: 1) blockade of both LFA-1 and Mac-1 further
decreased monocyte crawling fraction; and 2) monocyte crawling
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 2. Monocyte crawling patterns become neutrophil-like upon TNF-a activation. In separate experiments, monocytes and neutrophils were
stained for F4/80 and GR-1, respectively (1.5 mg/mouse, i.v.). A, The number of adhered cells/field of view. pSignificantly different from appropriate
control group (p , 0.05); ppp , 0.01. Time-lapsed microscopy (390) was used to track crawling leukocytes. B, The fraction of crawling leukocytes out of
total adhered. For A and B, n = 11 venules, four mice. The crawling velocities (C), the crawling distance (D), and the confinement ratio (E) (as defined in
Materials and Methods) were quantified. n = 84–125 leukocytes, 4 mice. ppSignificantly different from all other groups (p , 0.01). F, Representative
monocyte crawling trajectories (n = 10) from unstimulated (upper panel) and TNF-a–activated (lower panel) venules. All starting positions were aligned to
the same origin. Black arrow indicates flow direction. Axes indicate distance in micrometers. G, Images of time-lapse series over 30 min of real-time
acquisition were summed to display the paths (red dotted lines) of three crawling monocytes (upper panel) versus three crawling neutrophils (lower panel)
in representative unstimulated venules over 40-min time period. Whereas neutrophils maintain similar crawling patterns in both unstimulated and TNF-a–
activated venules, monocytes switch from long-distance, unconfined crawling to neutrophil-like, short-distance, and more direct crawling.
7062
distances under these conditions were similar to crawling distances of
monocytes and neutrophils in TNF-a–activated venules (67.4 6 9.6
and 61.8 6 5.4 mm, respectively; Fig. 2D), which are both Mac-1
dependent. Furthermore, Mac-1 block altered the characteristically
short-distance, neutrophil-like crawling of monocytes in TNF-a–
activated venules (67.4 6 9.6 mm; Fig. 2D) to be significantly longer
(147.6 6 15.2 mm, Fig. 4B); this length was similar to LFA-1–
mediated monocyte crawling distance in unstimulated venules.
Blockade of LFA-1 had no significant effect on monocyte crawling
distance (Fig. 4B). Consistent with these observations, in unstimulated venules in Mac-1 KO mice, the average monocyte crawling
distance was 138.2 6 11.5 mm, which was not significantly different
from monocyte crawling distances in unstimulated venules in WT
mice. Importantly, in TNF-a–activated venules, Mac-1 KO monocytes that remained crawling continued to crawl for long distances
(122.9 6 11.1 mm; Supplemental Fig. 6), similar to that reported for
both unstimulated venules in WT mice and in WT mice in the
presence of Mac-1–blocking Ab (Fig. 4). Together, these data confirm our findings (Figs. 3, 4) that monocytes can use both Mac-1 and
LFA-1 for crawling; however, LFA-1 characteristically mediates
longer crawling distances than Mac-1. Crawling for longer distances
is consistent with the idea that one of the important functions of
FIGURE 4. LFA-1–mediated crawling is for characteristically longer
distances than that mediated by Mac-1. In separate experiments, monocytes and neutrophils were stained for F4/80, and GR-1 respectively, (1.5
mg/mouse, i.v.) and were treated with nonspecific IgGs, Mac-1–blocking
Ab, LFA-1–blocking Ab, a combination of both Abs (50 mg/mouse, i.v.),
or VLA-4–blocking Ab. The effects of these blocking Abs on the crawling
distance and the crawling velocity of both monocytes and neutrophils were
quantified in unstimulated tissue (A, C) and tissue that was treated with
TNF-a 4 h prior to Ab administration (B, D). The effects of both IgG2a (a
control isotope for LFA-1) and IgG2b (control isotope for Mac-1 and
VLA-4) on leukocyte adhesion and crawling were tested. As no significant
differences were found between the two nonspecific IgGs for each condition, for simplicity, all data are presented against IgG2b. LFA-1 was able
to sustain significantly longer crawling distances than Mac-1. VLA-4 block
had no significant effect on the crawling distances of either monocytes or
neutrophils. For all groups, n = 11 venules, 4–6 mice. p/&Significantly
different from each other (p , 0.05); pp/^^p , 0.01.
monocytes is to survey the unstimulated venules. Mac-1–mediated
crawling distance is significantly shorter and is therefore likely to
increase the number of leukocytes that transmigrate during inflammation. There were no significant differences in the crawling
velocities of both monocytes and neutrophils in TNF-a–activated
venules (Fig. 4D). Intriguingly, we also found that the remaining
fraction of crawling neutrophils in TNF-a–activated venules following Mac-1 block (likely LFA-1 mediated) crawled for significantly longer distances compared with distances with control IgG
or in untreated venules activated with TNF-a (126.6 6 10.2 mm
Mac-1 block versus 61.6 6 8.1 mm IgG; Fig. 4B). This observation
was also confirmed in Mac-1 KO mice (Supplemental Fig. 6),
suggesting that in the absence of functional Mac-1, neutrophils are
also able to use LFA-1 for crawling and further supports our conclusion that LFA-1–mediated crawling is characteristically longer
than that mediated by Mac-1.
Expression of LFA-1 and Mac-1 on monocytes and neutrophils
in situ
We hypothesized that the interchangeable roles of LFA-1 and
Mac-1 in monocyte crawling would likely be reflected by changes
in the expression of LFA-1 and Mac-1. Using flow cytometry, we
measured the expression of LFA-1 and Mac-1 on monocytes and
neutrophils isolated from control and TNF-a–treated mouse circulations. As expected (37), the expression of Mac-1 on both
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 3. Roles for LFA-1, Mac-1, and VLA-4 in mediating monocyte
and neutrophil crawling in unstimulated and inflamed venules. In separate
experiments, monocytes and neutrophils were stained for F4/80 and GR-1,
respectively (1.5 mg/mouse, i.v.), and were treated with nonspecific IgGs,
Mac-1–blocking Ab, LFA-1–blocking Ab, a combination of both Abs (50
mg/mouse, i.v.), or VLA-4–blocking Ab. The effects of these blocking Abs
on the ability of both monocytes and neutrophils to adhere and crawl were
quantified in unstimulated tissue (A, C) and tissue that was treated with
TNF-a 4 h prior to Ab administration (B, D). The effects of both IgG2a (a
control isotope for LFA-1) and IgG2b (control isotope for Mac-1 and
VLA-4) on leukocyte adhesion and crawling were tested. As no significant
differences were found between the two nonspecific IgGs for each condition, for simplicity, all data are presented against IgG2b. Whereas neutrophil crawling was Mac-1 dependent in both unstimulated and TNF-a–
activated venules, monocytes primarily used LFA-1 in unstimulated and
Mac-1 in TNF-a–activated venules. VLA-4 block significantly attenuated
monocyte adhesion, but had no significant effect on either monocyte or
neutrophil crawling. For all groups, n = 11 venules, 4–6 mice. pSignificantly different from IgG treatment within the same treatment group
(p , 0.05); pp/^^p , 0.01; &significantly different from each other (p ,
0.01).
LEUKOCYTE CRAWLING AND EMIGRATION IN SITU
The Journal of Immunology
Transmigration of neutrophils and monocytes following TNF-a
activation
We showed that Mac-1 plays a more prominent role in mediating
both neutrophil and monocyte crawling in TNF-a–activated venules compared with LFA-1. Thus, next we asked whether Mac-1 is
also more critical for leukocyte TEM into extravascular sites under these conditions. To test this, we quantified leukocyte TEM in
cremaster venules and leukocyte emigration into the peritoneal
cavity in response to TNF-a activation. In noninflamed, surgically
prepared tissue surrounding the cremaster venules, transmigrated
leukocytes are rarely observed (16); however, TNF-a treatment resulted in a robust increase in the number of leukocytes in extravascular regions (15 6 1.0 leukocyte/10,000 mm2; Fig. 6A). We
found that whereas blocking Abs for both LFA-1 and Mac-1 (but
FIGURE 5. In situ administration of TNF-a increases the expression of
Mac-1 but not LFA-1 and VLA-4 on circulating monocytes and neutrophils. Flow cytometry was used to measure the surface expression of
Mac-1, LFA-1, and VLA-4 on monocytes and neutrophils isolated from
mouse circulation under unstimulated and TNF-a–activated conditions.
Isolated leukocytes were double-labeled with either anti-F4/80 or Gr-1 (10
mg/ml) in combination with CD11a, CD11b, or CD49d (20 mg/ml), and the
number of Ab binding sites per cell was established as defined in Materials
and Methods. Both monocytes and neutrophils isolated from TNF-a–
treated mice significantly increased Mac-1 expression; however, the expression of LFA-1 on monocytes was significantly lower compared with
unstimulated tissue, but remained unchanged on neutrophils. Both monocytes and neutrophils expressed VLA-4 (neutrophils at very low levels that
did not change following TNF-a treatment). For all groups, n = 3 mice.
ppSignificantly different from each other (p , 0.01).
not nonspecific IgG Abs) significantly decreased leukocyte transmigration (10 6 0.9 and 3.7 6 0.6 leukocyte/10,000 mm2, respectively; Fig. 6A), a more prominent effect was observed following Mac-1 block. As Mac-1 is essential for both monocyte and
leukocyte crawling in TNF-a–activated venules (Fig. 3), these
findings further support the idea that getting to the transmigratory
portals is critical for TEM. We also, in separate experiments,
evaluated the effect of LFA-1 and Mac-1 block on monocyte and
neutrophil TEM. Consistent with what we measured for the total
leukocyte population (Fig. 6A), block of Mac-1 was significantly
more efficient in blocking TEM of both monocytes and neutrophils (5.7 6 0.4 versus 1.9 6 0.2 monocytes/10,000 mm2 and
9.3 6 0.8 versus 1.9 6 0.4 neutrophils/10,000 mm2; Fig. 6A)
compared with block of LFA-1 (which also significantly decreased
TEM). The representative images in Fig. 6C show neutrophils
(stained for GR-1, upper panels) and monocytes (stained for F4/80,
bottom panels) within the blood vessels (as indicated by white lines)
or in the extravascular space under the conditions specified on each
panel. We further tested the effects of Mac-1– or LFA-1–blocking
Abs on the emigration of all leukocytes, and specifically neutrophils, into the peritoneal cavity. Under control conditions (data
not shown), the majority of leukocytes residing in the peritoneal
cavity are mononuclear cells. TNF-a treatment induced leukocyte
infiltration into the peritoneal cavity of WT animals (0.5 6 0.05
control versus 1.6 6 0.2 TNF-a, 3 106 leukocyte/3 ml volume of
lavage; Fig. 7A), primarily due to accumulation of neutrophils
(7.2 6 0.8% control versus 47.8 6 2.3% TNF-a; Fig. 7B). Consistent with the cremaster tissue observations, blockade of Mac-1
abrogated the TNF-a–induced leukocyte emigration into the peritoneal cavity (1.6 6 0.2 versus 0.47 6 0.03, 3 106 leukocyte/
cavity; Fig. 7A) and decreased the fraction of emigrated neutrophils to 25.3 6 2.1% of the emigrated cells (Fig. 7B). The decrease in leukocyte emigration following LFA-1 block was not
significant in this model (Fig. 7B).
Discussion
Sequential leukocyte rolling, adhesion, and TEM constitute a complex multistep paradigm that is orchestrated by multiple adhesion
molecules. Understanding of this sequence of events has been
expanded with the recent recognition of postadhesion leukocyte
spreading and intralumenal crawling (7) and the documentation
that this additional step is essential for TEM (24). Together with
the findings that regions enriched in ICAM-1 indicate portals for
leukocyte TEM (16), as do less dense regions of basal lamina (38),
and the possible role for ICAM-1 in guiding crawling leukocytes
toward these portals, this has offered expanded paradigms concerning leukocyte recruitment during inflammation. Thus, the recognition of leukocyte crawling has not only identified a mechanism for leukocyte motility within the blood vessels, but has also
revealed a physiological role for the heterogeneity of endothelial
morphology and molecular composition that makes some regions
better equipped to support leukocyte TEM than others.
The behavior of different leukocyte subtypes in response to inflammatory stimuli has been studied extensively, but less is known
about leukocyte–EC interactions in unstimulated (physiological)
conditions. Recent work has described monocyte adhesion and
crawling in unstimulated venules (15). Importantly, monocyte
crawling patterns under these conditions have been described as
exploratory and are thought to support the unique function of
monocytes to monitor healthy tissue. In this study, we extend these
findings by showing that this behavior is indeed characteristic
of monocytes but not neutrophils (Fig. 1). Moreover, monocyte
crawling in unstimulated venules is different from that in activated
vessels or from that of neutrophils, as it is for significantly longer
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
monocytes and neutrophils significantly increased following TNF-a
treatment (26.9 6 1.5 unstimulated versus 50.0 6 1.4 with TNF-a
on monocytes and 33.1 6 1.2 versus 56.7 6 0.8 on neutrophils, 3
104 molecules/cell; Fig. 5). The expression of LFA-1 on monocytes following TNF-a treatment was significantly lower than that
in unstimulated tissue (2.9 6 0.3 versus 9.9 6 0.4, 3 10 4
molecules/cell; Fig. 5), but remained unchanged on neutrophils
(9.6 6 0.2 TNF-a versus 8.4 6 0.1 control, 3 104 molecules/cell).
The increased expression of Mac-1 by monocytes (together with
the decreased expression of LFA-1) after TNF-a activation results
in a ∼1.5-fold increase in the total b2 integrin density on thesecells and indicates a mechanism whereby crawling that was LFA-1
mediated could become dependent on Mac-1, as described in Fig.
3, in TNF-a–activated venules. Additionally, as shown in Fig. 3,
VLA-4 block significantly decreased monocyte (but not neutrophil)
adhesion in inflamed venules; thus, we also measured the expression of VLA-1 on monocytes and neutrophils from unstimulated
and TNF-a–activated mice. Monocytes express VLA-4, and this
expression was not significantly different in unstimulated versus
activated conditions (6.9 6 0.7 and 6.8 6 0.7, 3 104 molecules/
cell; Fig. 5). Neutrophils expressed very low levels of VLA-4,
which did not change upon activation with TNF-a (Fig. 5).
7063
7064
LEUKOCYTE CRAWLING AND EMIGRATION IN SITU
distances and less direct (Fig. 2). It is also possible that the two
subsets of circulating monocytes, inflammatory, which are considered to specialize in proinflammatory activities, and resident,
FIGURE 7. Total leukocyte and specifically neutrophil emigration into
the peritoneal cavity is primarily Mac-1 dependent. Selected mice were
injected with saline (200 ml), Mac-1–blocking Ab, LFA-1–blocking Ab, or
nonspecific IgG Abs (50 mg/mouse in 200 ml saline) 10 min prior to the
injection of TNF-a (100 ng in 50 ml saline, ip.). Four hours later, peritoneal
cavities of anesthetized mice were lavaged, and white cells were recovered.
A, The total number of leukocytes was counted using a hemocytometer. B,
Neutrophil counts were determined on 100 ml cytospins stained with DiffQuik and presented as percent total population. For all groups in A and B,
n = three to four mice. The effects of both IgG2a (control isotope for LFA1) and IgG2b (control isotope for Mac-1) on leukocyte adhesion and
crawling were tested. As no significant differences were found between
the two nonspecific IgGs for each condition, for simplicity, all data are
presented against IgG2b. Consistent with the cremaster observations,
Mac-1 plays a more prominent role in leukocyte TEM compared with
LFA-1. pp/&Significantly different from each other (p , 0.01).
which play a role in tissue repair in microvessels, might use different crawling patterns to achieve these functions. Our data also
show that the distinct crawling patterns are a function of the
specific integrin mediating this crawling. We found that LFA-1–
dependent leukocyte crawling is sustained for longer distances and
greatly deviates from a straight line, whereas Mac-1–dependent
crawling distances are significantly shorter and more direct (Fig.
2). The longer and more meandering crawling distances coincide
with the surveying behavior (of monocytes), which has been identified as typical of unstimulated tissue, whereas the shorter crawling distances characterize both monocyte and neutrophil crawling
during inflammation. As the immune response requires rapid mobilization of both neutrophils and monocytes from the blood
vessels into the tissue during inflammation, crawling for short
distances and undertaking the most direct route toward emigration
will likely achieve this more efficiently and will result in a greater
number of emigrated leukocytes during a shorter time period.
Clearly different signaling pathways must be employed to sustain
prolonged versus shorter crawling distances, and it will be of great
importance to define these mechanisms in future studies.
Monocyte adhesion and crawling in unstimulated venules is
LFA-1 dependent (Fig. 3). LFA-1 is constitutively expressed by
both monocytes and neutrophils (Fig. 5); thus, it is puzzling why
monocytes are able to adhere and crawl in unstimulated venules,
but neutrophils are not. One possible mechanism is via differences
in the activation state of LFA-1 on monocytes and neutrophils.
Both LFA-1 and Mac-1 exist in different conformation-dependent
affinity states: inactive, intermediate, and active (39, 40). This dictates the ability to bind endothelial ligands, such as ICAM-1 (41),
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
FIGURE 6. Mac-1 plays a more prominent role than LFA-1 in leukocyte TEM in cremaster venules. Leukocyte TEM in cremaster tissue in response to
TNF-a treatment was quantified in the absence (2) or in the presence of a nonspecific IgG (data not shown) or LFA-1– or Mac-1–blocking Abs. Monocyte
and neutrophil markers F4/80 and GR-1, respectively (1.5 mg/mouse), as well as all Ab solutions (50 mg/mouse) were administered via tail vein injection 10
min prior to the injection of TNF-a (100 ng in 50 ml saline, i.p.) and 4 h prior to observations. A, The total number of all leukocytes transmigrated into the
extravascular tissue (100,000 mm2) was quantified using bright field microscopy. B, In separate experiments, transmigrated monocytes and neutrophils
stained for F4/80 and GR-1, respectively, were counted in the extravascular tissue (100,000 mm2) under the specified conditions using fluorescence illumination. For all groups, n = 10–13 venules, 4–6 mice. C, Representative images (Z-stack projection of microvessels confocally scanned from the upper
wall through the middle of the vessel) of neutrophils (upper panels) and monocytes (bottom panels) within the blood vessels (outlined with white lines) and
in the surrounding tissue under the conditions specified on each panel. Scale bar, 15 mm. Blockade of both LFA-1 and Mac-1 significantly reduced
leukocyte TEM, but Mac-1 block was more effective (as quantified in B). pSignificantly different from TNF-a alone (p , 0.05); ppsignificantly different
from TNF-a alone (p , 0.01).
The Journal of Immunology
surprising as: 1) activated monocytes and neutrophils express significantly higher levels of Mac-1 compared with LFA-1; and 2)
Mac-1 is predominately used by leukocytes to get to the specific
venular portals where TEM takes place. In TNF-a–activated venules following blockade of Mac-1, both monocytes and neutrophils
were still able to adhere to the endothelium; however, the majority
of leukocytes remained stationary until they detached and washed
away in the flowing blood. Due to the EC morphology, leukocytes
by default land on EC junctions (7, 16), which are known to accommodate most TEM (10). Thus, although it is possible that some
leukocytes landed on active portals (the specific location that is
equipped to accommodate leukocyte TEM), and others were able to
punch their way through the endothelium (transcellular route), as
has been previously suggested (48), the current study demonstrates
that intralumenal crawling is crucial for the majority of leukocyte
recruitment during inflammation.
Acknowledgments
We thank J.M. Kuebel and O.A. Spindel for technical contributions.
Disclosures
The authors have no financial conflicts of interest.
References
1. Smith, J. A. 1994. Neutrophils, host defense, and inflammation: a double-edged
sword. J. Leukoc. Biol. 56: 672–686.
2. Voisin, M. B., A. Woodfin, and S. Nourshargh. 2009. Monocytes and neutrophils
exhibit both distinct and common mechanisms in penetrating the vascular
basement membrane in vivo. Arterioscler. Thromb. Vasc. Biol. 29: 1193–1199.
3. Muller, W. A., and G. J. Randolph. 1999. Migration of leukocytes across endothelium and beyond: molecules involved in the transmigration and fate of
monocytes. J. Leukoc. Biol. 66: 698–704.
4. Kansas, G. S. 1996. Selectins and their ligands: current concepts and controversies. Blood 88: 3259–3287.
5. Ley, K., M. Allietta, D. C. Bullard, and S. Morgan. 1998. Importance of
E-selectin for firm leukocyte adhesion in vivo. Circ. Res. 83: 287–294.
6. Phillipson, M., B. Heit, P. Colarusso, L. Liu, C. M. Ballantyne, and P. Kubes.
2006. Intraluminal crawling of neutrophils to emigration sites: a molecularly
distinct process from adhesion in the recruitment cascade. J. Exp. Med. 203:
2569–2575.
7. Wojciechowski, J. C., and I. H. Sarelius. 2005. Preferential binding of leukocytes
to the endothelial junction region in venules in situ. Microcirculation 12: 349–
359.
8. McEver, R. P., and R. D. Cummings. 1997. Role of PSGL-1 binding to selectins
in leukocyte recruitment. J. Clin. Invest. 100(11 Suppl): S97–S103.
9. Steeber, D. A., M. L. Tang, N. E. Green, X. Q. Zhang, J. E. Sloane, and
T. F. Tedder. 1999. Leukocyte entry into sites of inflammation requires overlapping interactions between the L-selectin and ICAM-1 pathways. J. Immunol.
163: 2176–2186.
10. Ley, K., C. Laudanna, M. I. Cybulsky, and S. Nourshargh. 2007. Getting to the
site of inflammation: the leukocyte adhesion cascade updated. Nat. Rev. Immunol. 7: 678–689.
11. Xie, X., J. Raud, P. Hedqvist, and L. Lindbom. 1997. In vivo rolling and endothelial selectin binding of mononuclear leukocytes is distinct from that of
polymorphonuclear cells. Eur. J. Immunol. 27: 2935–2941.
12. Henderson, R. B., L. H. Lim, P. A. Tessier, F. N. Gavins, M. Mathies, M. Perretti,
and N. Hogg. 2001. The use of lymphocyte function-associated antigen (LFA)-1deficient mice to determine the role of LFA-1, Mac-1, and alpha4 integrin in the
inflammatory response of neutrophils. J. Exp. Med. 194: 219–226.
13. Meerschaert, J., and M. B. Furie. 1995. The adhesion molecules used by
monocytes for migration across endothelium include CD11a/CD18, CD11b/
CD18, and VLA-4 on monocytes and ICAM-1, VCAM-1, and other ligands
on endothelium. J. Immunol. 154: 4099–4112.
14. Luscinskas, F. W., G. S. Kansas, H. Ding, P. Pizcueta, B. E. Schleiffenbaum,
T. F. Tedder, and M. A. Gimbrone, Jr. 1994. Monocyte rolling, arrest and
spreading on IL-4-activated vascular endothelium under flow is mediated via
sequential action of L-selectin, beta 1-integrins, and beta 2-integrins. J. Cell Biol.
125: 1417–1427.
15. Auffray, C., D. Fogg, M. Garfa, G. Elain, O. Join-Lambert, S. Kayal, S. Sarnacki,
A. Cumano, G. Lauvau, and F. Geissmann. 2007. Monitoring of blood vessels
and tissues by a population of monocytes with patrolling behavior. Science 317:
666–670.
16. Sumagin, R., and I. H. Sarelius. 2010. Intercellular adhesion molecule-1 enrichment near tricellular endothelial junctions is preferentially associated with
leukocyte transmigration and signals for reorganization of these junctions to
accommodate leukocyte passage. J. Immunol. 184: 5242–5252.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
consequently affecting leukocyte–EC interactions. We speculate
that under unstimulated conditions, monocytes have a higher
number of activated LFA-1 than neutrophils; however, this remains
to be determined. An additional candidate likely contributing to
leukocyte–EC interactions is VLA-4 (CD49d/CD29). It is abundantly expressed on monocytes (13) (Fig. 5) and is used by monocytes to adhere to the inflamed endothelium via interaction with
VCAM-1 (Fig. 3), the expression of which significantly increases
following TNF-a treatment (35). VLA-4 was also detected at low
levels on neutrophils, but in our study, it played no significant role
in neutrophil adhesion to the inflamed endothelium (Fig. 4). A
previous in situ study showed a role for VLA-1 in neutrophil rolling
in WT mice, as well as in supporting leukocyte adhesion in the
absence of functional LFA-1 (12), suggesting that VLA-4 was able
to compensate for the loss of LFA-1. Not surprisingly, we found no
significant contribution of VLA-4 to both monocyte and neutrophil
adhesion in unstimulated microvessels, as we (35) and others (42)
have previously determined that there is very low expression of
VCAM-1 on resting endothelium in microvessels. Our finding that
some crawling monocytes remain following blockade of both
LFA-1 and Mac-1 (Fig. 5), together with the finding that VLA-4 is
not involved in monocyte crawling, suggests that additional factors
other than VLA-4 are involved in this process; this remains to be
determined in future work.
The expression of Mac-1 is known to increase upon leukocyte
activation (43), but the constitutively expressed LFA-1 remains
unchanged (44). We confirmed this by showing that activation of
isolated monocytes and neutrophils with TNF-a in vitro (4 h)
resulted in significant upregulation of Mac-1 expression; however,
the levels of LFA-1 remained unchanged (data not shown). Similarly, both monocytes and neutrophils that were isolated from TNFa–treated mice showed a dramatic increase in Mac-1 compared
with leukocytes isolated from unstimulated mice. Surprisingly,
monocytes isolated from TNF-a–treated mice showed decreased
levels of LFA-1 compared with monocytes from unstimulated mice
(Fig. 5). We speculate that the loss of LFA-1 might be a result of
LFA-1 internalization or shedding (45, 46) o, alternatively, a result
of direct competition between the aL and aM subunits, as both
LFA-1 and Mac-1 share the same b-chains (CD18 subunits). Despite the decrease in monocyte LFA-1, the much more substantial
increase in Mac-1 expression resulted in a ∼1.5-fold increase in the
total expression of surface b2 integrins following TNF-a treatment.
The same was true for neutrophils (Fig. 5). These observations
suggest that the dramatic increase in Mac-1 expression is the reason
that LFA-1–dependent monocyte crawling in resting tissue becomes mainly Mac-1 dependent upon TNF-a activation. Interestingly, the crawling velocities of monocytes and neutrophils in
either unstimulated or inflamed tissue are not different (Fig. 4B),
suggesting that the kinetics of bond formation between LFA-1 and
Mac-1 and their endothelial ligands, such as ICAM-1, are not rate
limiting for cell migration.
Finally, LFA-1 has been shown to play an important role in
leukocyte TEM (47). However, in our hands, and in agreement with
other recent work (6), Mac-1 is significantly more important for
this process. As shown in Figs. 6 and 7, blockade of Mac-1 was
significantly more effective than blockade of LFA-1 in reducing
leukocyte TEM in both cremaster venules and the peritoneal cavity.
We cannot eliminate the possibility that using a mAb against Mac-1
might trigger outside in signaling, thus potentially affecting leukocyte function and preventing TEM. However, it appears unlikely
that Mac-1 cross-linking would lead to an inhibition of leukocyte
function, because integrin ligation will lead to leukocyte activation,
resulting in increased adhesion and TEM. The finding that Mac-1 is
more important in regulating leukocyte TEM than LFA-1 is not
7065
7066
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
granulocyte-differentiation antigen (Gr-1) detects members of the Ly-6 family. J.
Immunol. 151: 2399–2408.
Ley, K., and P. Gaehtgens. 1991. Endothelial, not hemodynamic, differences are
responsible for preferential leukocyte rolling in rat mesenteric venules. Circ.
Res. 69: 1034–1041.
Sumagin, R., and I. H. Sarelius. 2007. A role for ICAM-1 in maintenance of
leukocyte-endothelial cell rolling interactions in inflamed arterioles. Am. J.
Physiol. Heart Circ. Physiol. 293: H2786–H2798.
Mackay, F., H. Loetscher, D. Stueber, G. Gehr, and W. Lesslauer. 1993. Tumor
necrosis factor alpha (TNF-alpha)-induced cell adhesion to human endothelial
cells is under dominant control of one TNF receptor type, TNF-R55. J. Exp.
Med. 177: 1277–1286.
Sengeløv, H., L. Kjeldsen, M. S. Diamond, T. A. Springer, and N. Borregaard.
1993. Subcellular localization and dynamics of Mac-1 (alpha m beta 2) in human
neutrophils. J. Clin. Invest. 92: 1467–1476.
Voisin, M. B., D. Pröbstl, and S. Nourshargh. 2010. Venular basement membranes ubiquitously express matrix protein low-expression regions: characterization in multiple tissues and remodeling during inflammation. Am. J. Pathol.
176: 482–495.
Shimaoka, M., M. Kim, E. H. Cohen, W. Yang, N. Astrof, D. Peer, A. Salas,
A. Ferrand, and T. A. Springer. 2006. AL-57, a ligand-mimetic antibody to
integrin LFA-1, reveals chemokine-induced affinity up-regulation in lymphocytes. Proc. Natl. Acad. Sci. USA 103: 13991–13996.
Luo, B. H., C. V. Carman, and T. A. Springer. 2007. Structural basis of integrin
regulation and signaling. Annu. Rev. Immunol. 25: 619–647.
Sarantos, M. R., S. Raychaudhuri, A. F. Lum, D. E. Staunton, and S. I. Simon.
2005. Leukocyte function-associated antigen 1-mediated adhesion stability is
dynamically regulated through affinity and valency during bond formation with
intercellular adhesion molecule-1. J. Biol. Chem. 280: 28290–28298.
Ley, K., and Y. Huo. 2001. VCAM-1 is critical in atherosclerosis. J. Clin. Invest.
107: 1209–1210.
Jacobson, P. B., and D. J. Schrier. 1993. Regulation of CD11b/CD18 expression
in human neutrophils by phospholipase A2. J. Immunol. 151: 5639–5652.
Crucian, B., M. Nelman-Gonzalez, and C. Sams. 2006. Rapid flow cytometry
method for quantitation of LFA-1-adhesive T cells. Clin. Vaccine Immunol. 13:
403–408.
Fukuda, S., and G. W. Schmid-Schönbein. 2003. Regulation of CD18 expression
on neutrophils in response to fluid shear stress. Proc. Natl. Acad. Sci. USA 100:
13152–13157.
Anderson, M. E., and T. J. Siahaan. 2003. Mechanism of binding and internalization of ICAM-1-derived cyclic peptides by LFA-1 on the surface of
T cells: a potential method for targeted drug delivery. Pharm. Res. 20: 1523–
1532.
Ding, Z. M., J. E. Babensee, S. I. Simon, H. Lu, J. L. Perrard, D. C. Bullard,
X. Y. Dai, S. K. Bromley, M. L. Dustin, M. L. Entman, et al. 1999. Relative
contribution of LFA-1 and Mac-1 to neutrophil adhesion and migration. J.
Immunol. 163: 5029–5038.
Carman, C. V., and T. A. Springer. 2004. A transmigratory cup in leukocyte
diapedesis both through individual vascular endothelial cells and between them.
J. Cell Biol. 167: 377–388.
Downloaded from http://www.jimmunol.org/ by guest on June 17, 2017
17. Wang, S., M. B. Voisin, K. Y. Larbi, J. Dangerfield, C. Scheiermann, M. Tran,
P. H. Maxwell, L. Sorokin, and S. Nourshargh. 2006. Venular basement membranes contain specific matrix protein low expression regions that act as exit
points for emigrating neutrophils. J. Exp. Med. 203: 1519–1532.
18. Soehnlein, O., A. Zernecke, E. E. Eriksson, A. G. Rothfuchs, C. T. Pham,
H. Herwald, K. Bidzhekov, M. E. Rottenberg, C. Weber, and L. Lindbom. 2008.
Neutrophil secretion products pave the way for inflammatory monocytes. Blood
112: 1461–1471.
19. Soehnlein, O., L. Lindbom, and C. Weber. 2009. Mechanisms underlying
neutrophil-mediated monocyte recruitment. Blood 114: 4613–4623.
20. Issekutz, A. C., and T. B. Issekutz. 1992. The contribution of LFA-1 (CD11a/
CD18) and MAC-1 (CD11b/CD18) to the in vivo migration of polymorphonuclear leucocytes to inflammatory reactions in the rat. Immunology 76: 655–
661.
21. Luu, N. T., G. E. Rainger, and G. B. Nash. 1999. Kinetics of the different steps
during neutrophil migration through cultured endothelial monolayers treated
with tumour necrosis factor-alpha. J. Vasc. Res. 36: 477–485.
22. Phillipson, M., B. Heit, S. A. Parsons, B. Petri, S. C. Mullaly, P. Colarusso,
R. M. Gower, G. Neely, S. I. Simon, and P. Kubes. 2009. Vav1 is essential for
mechanotactic crawling and migration of neutrophils out of the inflamed microvasculature. J. Immunol. 182: 6870–6878.
23. Bradfield, P. F., C. Scheiermann, S. Nourshargh, C. Ody, F. W. Luscinskas,
G. E. Rainger, G. B. Nash, M. Miljkovic-Licina, M. Aurrand-Lions, and
B. A. Imhof. 2007. JAM-C regulates unidirectional monocyte transendothelial
migration in inflammation. Blood 110: 2545–2555.
24. Schenkel, A. R., Z. Mamdouh, and W. A. Muller. 2004. Locomotion of monocytes on endothelium is a critical step during extravasation. Nat. Immunol. 5:
393–400.
25. Sumagin, R., and I. H. Sarelius. 2006. TNF-alpha activation of arterioles and
venules alters distribution and levels of ICAM-1 and affects leukocyteendothelial cell interactions. Am. J. Physiol. Heart Circ. Physiol. 291: H2116–
H2125.
26. Sarelius, I. H. 1993. Cell and oxygen flow in arterioles controlling capillary
perfusion. Am. J. Physiol. 265: H1682–H1687.
27. Sumagin, R., E. Lomakina, and I. H. Sarelius. 2008. Leukocyte-endothelial cell
interactions are linked to vascular permeability via ICAM-1-mediated signaling.
Am. J. Physiol. Heart Circ. Physiol. 295: H969–H977.
28. Chiang, E. Y., A. Hidalgo, J. Chang, and P. S. Frenette. 2007. Imaging receptor
microdomains on leukocyte subsets in live mice. Nat. Methods 4: 219–222.
29. Kaynar, A. M., A. M. Houghton, E. H. Lum, B. R. Pitt, and S. D. Shapiro. 2008.
Neutrophil elastase is needed for neutrophil emigration into lungs in ventilatorinduced lung injury. Am. J. Respir. Cell Mol. Biol. 39: 53–60.
30. Bixel, M. G., B. Petri, A. G. Khandoga, A. Khandoga, K. Wolburg-Buchholz,
H. Wolburg, S. März, F. Krombach, and D. Vestweber. 2007. A CD99-related
antigen on endothelial cells mediates neutrophil but not lymphocyte extravasation in vivo. Blood 109: 5327–5336.
31. Lagasse, E., and I. L. Weissman. 1996. Flow cytometric identification of murine
neutrophils and monocytes. J. Immunol. Methods 197: 139–150.
32. Geissmann, F., S. Jung, and D. R. Littman. 2003. Blood monocytes consist of
two principal subsets with distinct migratory properties. Immunity 19: 71–82.
33. Fleming, T. J., M. L. Fleming, and T. R. Malek. 1993. Selective expression of
Ly-6G on myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb to
LEUKOCYTE CRAWLING AND EMIGRATION IN SITU