Am J Physiol Cell Physiol
280: C814–C822, 2001.
Human neutrophils facilitate tumor cell
transendothelial migration
QIONG DI WU,1 JIANG HUAI WANG,1 CLAIRE CONDRON,2
DAVID BOUCHIER-HAYES,2 AND H. PAUL REDMOND1
Department of Surgery, 1Cork University Hospital, University College Cork, Cork,
and 2Beaumont Hospital, Royal College of Surgeons in Ireland, Dublin, Ireland
Received 27 March 2000; accepted in final form 9 November 2000
metastasis; endothelial cells; adhesion molecules; cell proliferation; apoptosis
DESPITE THE ADVANCES in primary tumor management,
up to 50% of patients will ultimately die of their disease. Metastases from primary tumors remain a major
cause of cancer-related deaths. To metastasize, tumor
cells must shed into the blood stream (intravasation)
directly by invasion into the tumor-derived vasculature
or indirectly by lymphatic drainage, survive in the
circulation, and finally migrate through normal vascular endothelium and proliferate in the target organs
(extravasation). Tumor cell extravasation plays a key
role in tumor metastasis. Several hypotheses have
been proposed to explain tumor cell extravasation (5, 7,
11). However, the precise mechanisms by which tumor
Address for reprint requests and other correspondence: H. P.
Redmond, Dept. of Surgery, Cork Univ. Hospital, Univ. College Cork,
Wilton, Cork, Ireland (E-mail: [email protected]).
C814
cells migrate through normal vascular endothelium
remain controversial.
The immune system acts to eliminate tumor cells.
However, under certain circumstances, leukocytes may
enhance the ability of weakly metastatic tumor cells to
metastasize, or tumor cells can exploit leukocyte function to increase their metastatic efficiency (1, 22). Human polymorphonuclear neutrophils (PMN), which
comprise 50–70% of circulating leukocytes, can be
spontaneously cytotoxic to tumor cells. However, PMN
may function to promote tumor growth and metastasis
rather than to cause tumor cytolysis. For example,
evidence from animal studies has shown that, in vitro,
PMN increases tumor cell attachment to endothelial
monolayers (21) and that tumor-elicited PMN, in contrast to normal PMN, enhances metastatic potential
and invasiveness of tumor cells in an in vivo tumorbearing rat model (24). Furthermore, under light and
electron microscopes, circulating PMN have been seen
in close association with metastatic tumor cells
throughout the process of tumor cell arrest and extravasation in vivo (4).
Human PMN actively participate in the inflammatory response by transendothelial migration. The process of PMN transmigration through vascular endothelium can be divided into three steps: PMN rolling,
adhesion, and migration. Each step is mediated by
different adhesion molecules (6, 17). PMN rolling is
mediated by L-selectin expressed on PMN and by Eselectin and P-selectin expressed on endothelial cells.
More importantly, both PMN adhesion and migration
require the expression of CD11a/CD18 (known as
LFA-1) and CD11b/CD18 (known as MAC-1) on PMN
and the expression of intercellular adhesion molecule-1
(ICAM-1) on endothelial cells. However, it is unknown
whether PMN transendothelial migration can be employed by tumor cells to assist their extravasation
during the process of metastasis.
In this study we test the hypotheses that normal
human PMN function is altered by tumor cells and that
these PMN can assist tumor cell migration through the
endothelial barrier. Using an in vitro transendothelial
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
0363-6143/01 $5.00 Copyright © 2001 the American Physiological Society
http://www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
Wu, Qiong Di, Jiang Huai Wang, Claire Condron,
David Bouchier-Hayes, and H. Paul Redmond. Human
neutrophils facilitate tumor cell transendothelial migration.
Am J Physiol Cell Physiol 280: C814–C822, 2001.—Tumor
cell extravasation plays a key role in tumor metastasis.
However, the precise mechanisms by which tumor cells migrate through normal vascular endothelium remain unclear.
In this study, using an in vitro transendothelial migration
model, we show that human polymorphonuclear neutrophils
(PMN) assist the human breast tumor cell line MDA-MB-231
to cross the endothelial barrier. We found that tumor-conditioned medium (TCM) downregulated PMN cytocidal function, delayed PMN apoptosis, and concomitantly upregulated
PMN adhesion molecule expression. These PMN treated with
TCM attached to tumor cells and facilitated tumor cell migration through different endothelial monolayers. In contrast, MDA-MB-231 cells alone did not transmigrate. FACScan analysis revealed that these tumor cells expressed high
levels of intercellular adhesion molecule-1 (ICAM-1) but did
not express CD11a, CD11b, or CD18. Blockage of CD11b and
CD18 on PMN and of ICAM-1 on MDA-MB-231 cells significantly attenuated TCM-treated, PMN-mediated tumor cell
migration. These tumor cells still possessed the ability to
proliferate after PMN-assisted transmigration. These results
indicate that TCM-treated PMN may serve as a carrier to
assist tumor cell transendothelial migration and suggest that
tumor cells can exploit PMN and alter their function to
facilitate their extravasation.
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
migration model, we demonstrate that the metastatic
human breast tumor cell line MDA-MB-231 cells fail to
transmigrate spontaneously. However, tumor-conditioned medium (TCM)-treated human PMN with resultant suppressed cytocidal function, delayed apoptosis, and increased adhesion receptor expression
facilitate the MDA-MB-231 cells to cross normal human macro- and microvascular endothelial monolayers. These results suggest that tumor cells can exploit
PMN function to enhance their metastatic potential at
the step of extravasation.
MATERIALS AND METHODS
231, isolated from a pleural effusion, was obtained from the
American Type Culture Collection (ATCC, Rockville, MD).
MDA-MB-231 cells were grown in L-15 medium supplemented with 10% FCS, penicillin (100 units/ml), streptomycin sulfate (100 ␮g/ml), and 2.0 mM glutamine. Cells were
maintained at 37°C in a humidified 5% CO2 atmosphere and
subcultured by trypsinization with 0.05% trypsin-0.02%
EDTA when cells became confluent.
TCM from MDA-MB-231 cells was prepared as follows.
MDA-MB-231 cells were grown to subconfluency (⬃80%) in
culture medium. After being washed three times with HBSS,
cells were incubated in fresh culture medium at 37°C in 5%
CO2 conditions for 24 h. TCM was then harvested, centrifuged at 700 g and 4°C for 20 min, passed through 0.22-␮m
pore-size filters (Gelman Sciences, Ann Arbor, MI), and
stored at ⫺20°C until use.
Assessment of neutrophil and MDA-MB-231 cell migration.
HUVEC and HMVEC in 0.5 ml of complete culture medium
(4 ⫻ 105 cells/ml) were grown in the upper chamber of a
12-mm-diameter, collagen-coated polytetrafluoroethylene
membrane Transwell (Costar, Cambridge, MA), with 1 ml of
complete culture medium in the lower chamber at 37°C in a
humidified 5% CO2 atmosphere for 30 h until confluent
monolayers were reached. The upper and lower chambers of
the Transwells were then washed twice with HBSS.
For the PMN-assisted MDA-MB-231 cell migration assay,
isolated PMN (1 ⫻ 106 cells/ml) were incubated with either
TCM or culture medium for 60 min. TCM-treated or mediumtreated PMN (5 ⫻ 104 cells) were mixed with MDA-MB-231
cells (2.5 ⫻ 104 cells) at a 2:1 ratio in a total volume of 0.5 ml
and added to the upper chamber above the endothelial cell
monolayers and collagen-coated polytetrafluoroethylene
membrane. One milliliter of fresh complete culture medium
with or without interleukin-8 (IL-8) (75 ng/ml, final concentration) was added to the lower chamber. After incubation at
37°C in a humidified 5% CO2 atmosphere for different time
points, migrated tumor cells and PMN in the lower chamber
were collected and counted according to their morphological
differences. In some experiments, PMN and tumor cells were
pretreated with saturating concentrations of different blocking MAbs for 30 min to identify the involvement of individual
integrin and ICAM-1 receptor in PMN-mediated tumor cell
migration.
Assessment of endothelial monolayer permeability. The upper chamber and the lower chamber of the Transwell were
washed three times with HBSS to remove unattached and
transmigrated cells 24 h after PMN-mediated MDA-MB-231
tumor cell migration. FITC-conjugated dextran was then
administered to the upper chamber at a concentration of 1
mg/ml. The medium of the upper and lower chambers was
aspirated separately after 4 h of incubation at 37°C in a
humidified 5% CO2 atmosphere. The fluorescence activity
was measured under excitation at 490 nm and emission at
530 nm. Pure FITC-dextran was used to produced a standard
reference curve. A permeability index was calculated according to the following formula: permeability index ⫽ [(experimental permeability ⫺ spontaneous permeability)/(membrane permeability ⫺ spontaneous permeability)] ⫻ 100%
Measurement of human PMN respiratory burst and phagocytosis. PMN respiratory burst was assessed using a BURSTTEST (Orpegen, Heidelberg, Germany). Briefly, 20 ␮l of the
fluorogenic substrate dihydrorhodamin 123 was added to 100
␮l of PMN suspension (1 ⫻ 106 cells/ml) and incubated at
37°C for 10 min. After centrifugation, the cell pellets were
resuspended in 100 ␮l of DNA-staining solution and incubated at 4°C for 15 min. A PHAGOTEST (Orpegen) was used
to determine PMN phagocytosis. Briefly, 20 ␮l of stabilized
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
Reagents and monoclonal antibodies. DMEM, M199 medium, L-15 medium, Hanks’ balanced salt solution (HBSS),
PBS without Ca2⫹ and Mg2⫹, FCS, penicillin, streptomycin
sulfate, fungizone, glutamine, and 0.05% trypsin-0.02%
EDTA solution were purchased from GIBCO BRL (Paisley,
UK). Endothelial cell growth supplement, 2% gelatin, Dglucose, Dextran, Tris, EDTA, sodium citrate, Triton X-100,
propidium iodide (PI), FITC-conjugated dextran (70 kDa),
and heparin were purchased from Sigma (St. Louis, MO).
Collagenase (type I) and Ficoll were obtained from Worthington (Freehold, NJ) and Pharmacia (Uppsala, Sweden), respectively. Mouse anti-human CD11a, CD11b, CD18, CD29,
CD61, and ICAM-1 monoclonal antibodies (MAbs) were purchased from Becton Dickinson (Mountain View, CA), Chemicon (Temecula, CA), and Serotec (Oxford, UK). Different
integrin and ICAM-1-blocking MAbs 4B4, ICRF44, TS1/18,
and P2A4 were purchased from Coulter Clone (Miami, FL),
Endogen (Woburn, MA), and Chemicon, respectively.
Endothelial cell culture. Human umbilical vein endothelial
cells (HUVEC) were isolated by collagenase treatment of
umbilical vein and cultured on 2% gelatin-coated culture
flasks (Falcon, Lincoln Park, NJ) in complete M199 medium
supplemented with 20% FCS, penicillin (100 U/ml), streptomycin sulfate (100 ␮g/ml), fungizone (0.25 ␮g/ml), heparin
(16 U/ml), endothelial cell growth supplement (75 ␮g/ml),
and 2.0 mM glutamine as previously described (10). Cells
were grown at 37°C in a humidified 5% CO2 condition and
subcultured by trypsinization with 0.05% trypsin-0.02%
EDTA when confluent monolayers were reached. Using the
immunofluorescence technique, we identified endothelial
cells by typical phase contrast “cobblestone” morphology and
by the presence of von Willebrand factor antigen. HUVEC
were used between passages 3 and 5.
Human dermal microvascular endothelial cells (HMVEC)
at passage 3 were purchased from Clonetics (San Diego, CA)
and cultured in endothelial cell basal medium (Clonetics)
supplemented with Bulletkit (Clonetics) including 5% FCS,
human epidermal growth factor (10 ng/ml), bovine brain
extract (12 ␮g/ml), hydrocortisone (1.0 ␮g/ml), gentamycin
(50 ␮g/ml), and amphotericin B (50 ng/ml). Cells were grown
at 37°C in a humidified 5% CO2 condition and split once a
week. HMVEC were used until passage 8.
Isolation of human PMN. Using the dextran-Ficoll technique as described previously (23), we isolated human PMN
from lithium-heparin anticoagulated blood collected from
healthy adult volunteers. Isolated PMN were resuspended in
complete DMEM containing 10% FCS, penicillin (100 U/ml),
streptomycin sulfate (100 ␮g/ml), and 2.0 mM glutamine.
PMN purity was ⬎98% with 99% viability as determined by
trypan blue exclusion.
Breast tumor cell line and preparation of TCM. The metastatic human breast adenocarcinoma cell line MDA-MB-
C815
C816
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
and opsonized FITC-labeled Escherichia coli suspension was
added to 100 ␮l of PMN suspension (1 ⫻ 106 cells/ml) and
incubated at 37°C for 10 min. After the addition of 100 ␮l of
quenching solution, the sample was washed twice with 3 ml
washing solution in an ice-water bath. After centrifugation,
100 ␮l of DNA-staining solution were added to cell pellets
and incubated for 15 min in an ice-water bath. The analysis
of PMN respiratory burst and phagocytosis was performed on
a FACScan flow cytometer (Becton Dickinson). The mean
channel fluorescence was detected with FL1 using logarithmic amplification on the basis of a minimum number of
10,000 cells collected and analyzed with the software Lysis
II.
Fluorescence-activated cell-sorter analysis of immunofluorescence. The expression of CD11a, CD11b, and CD18 on
human PMN and the expression of ␤1-, ␤2-, and ␤3-integrins
and ICAM-1 on MDA-MB-231 cells were assessed by the
addition of 20 ␮l of FITC-conjugated anti-LFA-1␣ (anti-
Fig. 2. Migrated MDA-MB-231 cells and human PMN stained using
Diff-Quick. Large cells are MDA-MB-231 cells, and cells with polymorphonuclears are human PMN. Note the attachment of PMN to
tumor cells.
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
Fig. 1. Tumor-conditioned medium (TCM)-treated human polymorphonuclear neutrophils (PMN) facilitate tumor cell transendothelial
migration. MDA-MB-231 cell transmigration through human umbilical vein endothelial cell (HUVEC) (A) and human dermal microvascular endothelial cell (HMVEC) (B) monolayers after 6 h of incubation was assessed as described in MATERIALS AND METHODS. MDA (1)
and MDA (2) represent 2.5 ⫻ 104 and 7.5 ⫻ 104 MDA-MB-231 cells.
Data are expressed as the means ⫾ SD and are representative of 5
separate experiments. The statistical significance was compared
with culture medium-treated PMN, *P ⬍ 0.05.
CD11a), phycoerythrin (PE)-conjugated anti-Leu-15 (antiCD11b), FITC-conjugated anti-LFA-1␤ (anti-CD18), PE-conjugated anti-CD29 (anti-␤1), FITC-conjugated anti-CD61
(anti-␤3), and PE-conjugated anti-Leu-54 (anti-ICAM-1)
MAbs to 100 ␮l of PMN or MDA-MB-231 cell suspension (1 ⫻
106 cells/ml). FITC- and PE-conjugated isotype IgG1 and
IgG2a MAbs were used as negative control. After incubation
at 4°C for 30 min, ␤1-, ␤2-, and ␤3-integrins and ICAM-1
expression on PMN and MDA-MB-231 cells were analyzed on
a FACScan flow cytometer for detecting the log of the mean
channel fluorescence intensity with an acquisition of FL1 and
FL2, respectively. The minimum number of 10,000 events
was collected and analyzed with the software Lysis II.
Assessment of apoptosis. TCM-treated, culture mediumtreated, and migrated PMN were incubated in 17 ⫻ 100-mm
polypropylene tubes (Falcon) at 37°C in a humidified 5% CO2
atmosphere for different times. PMN apoptosis was assessed
according to the percentage of cells with hypodiploid DNA
using the PI staining technique (13). Briefly, after centrifugation, 1 ⫻ 106 cells were gently resuspended in 1 ml of
hypotonic fluorochrome solution (50 ␮g/ml PI, 3.4 mM sodium citrate, 1.0 mM Tris, 0.1 mM EDTA, and 0.1% Triton
X-100) and incubated in the dark at 4°C for 2 h before they
were analyzed with a FACScan flow cytometer (Becton Dickinson). The forward scatter and side scatter of cell particles
were simultaneously measured. The PI fluorescence of individual nuclei with an acquisition of FL2 was plotted against
forward scatter, and the data were registered on a logarithmic scale. The minimum number of 5,000 events was collected and analyzed with the software Lysis II. Apoptotic cell
nuclei were distinguished by their hypodiploid DNA content
from the diploid DNA content of normal cell nuclei. Cell
debris was excluded from analysis by raising the forward
threshold. All measurements were performed under the
same instrument settings.
Determination of cell proliferation. Cell proliferation was
determined using 5-bromo-2⬘-deoxy-uridine (BrdU) labeling
and detection ELISA kit (Boehringer Mannheim, Mannheim,
Germany). Briefly, 100 ␮l of control and migrated MDA-MB231 cell suspension (7.5 ⫻ 104 cells/ml) were seeded in 96well plates (Costar) and incubated at 37°C in a humidified 5%
CO2 atmosphere for 24 h. BrdU (10 ␮M) was added to culture
medium and incorporated into freshly synthesized DNA.
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
Fixed cells with DNA partially digested by nucleases were
incubated with anti-BrdU MAb conjugated with peroxidase
to detect incorporated BrdU. The absorbance of the colored
reaction product, which is directly correlated to the level of
BrdU incorporated into cellular DNA, was determined on a
Microtiter Plate Reader (Dynex Technologies, Chantilly, VA).
Statistical analysis. All data are presented as the means ⫾
SD. Statistical analysis was performed using ANOVA. Differences were judged statistically significant when the P
value was ⬍0.05.
RESULTS
Promotion of tumor cell transendothelial migration
by TCM-treated PMN. MDA-MB-231 cells at different
concentrations (2.5 ⫻ 104 and 7.5 ⫻ 104 cells) were
cocultured with endothelial cell monolayers to assess
tumor cell spontaneous migration. As shown in Fig. 1,
A and B, tumor cells themselves did not migrate spon-
Fig. 4. Endothelial monolayer permeability 24 h after human PMN
and MDA-MB-231 cell transmigration. Permeability index after 4 h
of incubation of FITC-conjugated dextran was assessed as described
in MATERIALS AND METHODS. MDA (1) and MDA (2) represent 2.5 ⫻ 104
and 7.5 ⫻ 104 MDA-MB-231 cells. Data are expressed as the
means ⫾ SD from 4 separate experiments. The statistical significance was compared with MDA-MB-231 cell alone, *P ⬍ 0.05.
Fig. 5. The effect of interleukin-8 (IL-8) on TCM-treated or culture
medium-treated PMN-induced tumor cell migration. MDA-MB-231
cell transmigration through endothelial monolayers following 6 h of
incubation was assessed as described in MATERIALS AND METHODS.
Data are expressed as the means ⫾ SD and are representative of 4
separate experiments. The statistical significance was compared
with culture medium-treated PMN, *P ⬍ 0.05, and TCM-treated
PMN without IL-8, @P ⬍ 0.05.
taneously through HUVEC and HMVEC monolayers.
Human PMN incubated with culture medium failed to
assist tumor cell transendothelial migration. However,
PMN treated with TCM significantly promoted MDAMB-231 cells crossing macro- and microvascular endothelial monolayers after 6 h of incubation (10.9 ⫾ 2.8
and 12.8 ⫾ 2.7% of tumor cell migration, respectively;
P ⬍ 0.05 vs. PMN incubated with culture medium).
Figure 2 shows migrated MDA-MB-231 cells after
TCM-treated PMN coculture and the attachment of
PMN to tumor cells. Maximum tumor cell migration
occurred at 6 h in a 24-h period of incubation, whereas
PMN treated with medium alone had no effect on
tumor cell migration (Fig. 3).
Fig. 6. MDA-MB-231 cell viability (A) and proliferation (B) before
and after transendothelial migration. Cell viability was determined
by trypan blue exclusion, and cell proliferation was measured as
described in MATERIALS AND METHODS. Data are expressed as the
means ⫾ SD from 5 separate experiments.
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
Fig. 3. The effect of TCM-treated or culture medium-treated human
PMN on tumor cell transmigration. MDA-MB-231 cell migration at
different time points was assessed as described in MATERIALS AND
METHODS. The results are expressed as the means ⫾ SD from 5
separate experiments.
C817
C818
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
Fig. 7. A: FACScan analysis of MDA-MB-231 cells
stained with FITC- or PE-conjugated negative monoclonal antibodies (MAbs, gray areas) and anti-CD11a,
-CD11b, -CD18, -ICAM-1, -␤1-integrin, and -␤3-integrin
MAbs (black areas) on the basis of a minimum number
of 10,000 cells collected. B: illustration of different
adhesion receptor expression on MDA-MB-231 cells.
The results are presented as log mean channel fluorescence per cell (MCF/cell). ICAM-1, intercellular adhesion molecule-1; data are the means ⫾ SD from 5
separate experiments.
To investigate whether TCM-treated PMN could
damage the endothelial monolayer, we performed an
endothelial monolayer permeability assay. As shown in
Fig. 4, TCM-treated PMN did not cause a significant
increase in endothelial monolayer permeability compared with culture medium-treated PMN, although the
permeability index in both culture medium-treated
PMN and TCM-treated PMN groups was significantly
higher than in MDA-MB-231 cell alone groups.
IL-8 is a potent chemotaxin and has been found to
increase PMN transmigration both in vitro and in vivo
(20). Therefore, we also assessed the effect of IL-8 on
PMN-induced tumor cell migration. IL-8 at 75 ng/ml
significantly enhanced MDA-MB-231 cell migration in
TCM-treated PMN group (Fig. 5). However, IL-8 did
not increase MDA-MB-231 cell migration in culture
medium-treated PMN group, although IL-8 significantly increased PMN themselves transmigration
(data not shown).
In vitro viability and proliferation of migrated tumor
cells. To evaluate the proliferative potential of migrated tumor cells, we assessed both viability and
proliferation of migrated MDA-MB-231 cells collected
after 6 h incubation. As shown in Fig. 6A, there was no
significant difference in viability between migrated
and control tumor cells (96 ⫾ 6 vs. 98 ⫾ 4%). Further-
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
Fig. 8. Upregulation of PMN adhesion receptor expression by TCM.
Normal human PMN were incubated with TCM or culture medium
for 60 min. PMN CD11a, CD11b, and CD18 expressions were assessed as described in MATERIALS AND METHODS and are presented as
log MCF/cell. Data are expressed as the means ⫾ SD and are
representative of 5 separate experiments. The statistical significance
was compared with culture medium-treated PMN, *P ⬍ 0.05.
Fig. 9. Inhibition of PMN respiratory burst and phagocytosis by
TCM. Normal human PMN were incubated with TCM or culture
medium for 60 min. PMN respiratory burst and phagocytosis were
assessed as described in MATERIALS AND METHODS and are presented
as log MCF/cell. Data are the means ⫾ SD from 5 separate experiments. The statistical significance was compared with culture medium-treated PMN, *P ⬍ 0.05.
rates at 6, 12, and 18 h compared with normal PMN
(P ⬍ 0.05).
The effect of blocking ␤-integrins and ICAM-1 on
TCM-treated, PMN-mediated tumor transmigration.
Different blocking MAbs were used to identify whether
the increased expression of ␤2-integrin (CD11b and
CD18) on PMN after TCM treatment and the constitutive expression of ␤1-integrin and ICAM-1 on MDAMB-231 tumor cells are involved in tumor cell transmigration. CD11b blocking MAb ICRF44, CD18
blocking MAb TS1/18, and ICAM-1 blocking MAb P2A4
significantly attenuated TCM-treated PMN-mediated
tumor cell migration (P ⬍ 0.05) (Fig. 11). Furthermore,
the combination of ICRF44 and P2A4 or TS1/18 and
P2A4 resulted in further reduction of tumor cell transmigration. However, ␤1-integrin blocking MAb 4B4
Fig. 10. The effect of TCM treatment on PMN apoptosis. Normal
human PMN were incubated with TCM or culture medium for 60
min. Migrated PMN were collected at 6 h after transendothelial
migration. PMN apoptosis was assessed at different time points by
propidium iodide (PI) DNA staining as described in MATERIALS AND
METHODS. Data are expressed as the means ⫾ SD from 5 separate
experiments. The statistical significance was compared with culture
medium-treated PMN, *P ⬍ 0.05.
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
more, the process of TCM-treated, PMN-induced tumor
cell transmigration did not impair the potential of
tumor cell proliferation, as migrated MDA-MB-231
cells still possessed 106 ⫾ 8% proliferation rate compared with the proliferation rate in control tumor cells
(Fig. 6B).
Expression of ␤-integrins and ICAM-1 on MDA-MB231 cells. To identify why MDA-MB-231 cells themselves did not migrate spontaneously through normal
human macro- and microvascular endothelial monolayers, we assessed ␤-integrin and ICAM-1 expression
on MDA-MB-231 cells. Figure 7A shows FACScan
analysis of MDA-MB-231 cells stained with control
MAbs and different adhesion molecule MAbs, and Fig.
7B illustrates ␤1-, ␤2-, ␤3-integrin and ICAM-1 expression on MDA-MB-231 cells. These tumor cells expressed high levels of ICAM-1 but did not express
CD11a/CD18 and CD11b/CD18, two adhesion receptors essential for adhesion to vascular endothelium
and migration through the endothelial barrier (6, 19).
MDA-MB-231 cells also contained ␤1-integrin, but not
␤3-integrin (Fig. 7, A and B).
TCM-induced alteration in PMN cytocidal function,
receptor expression, and PMN apoptosis. To elucidate
the possible mechanisms by which TCM-treated PMN
facilitated tumor cell transmigration, we measured
PMN adhesion receptor expression and cytocidal function after TCM treatment. Exposure of normal human
PMN to TCM for 60 min significantly enhanced CD11b
and CD18 receptor expression (P ⬍ 0.05 vs. PMN
incubated with culture medium) (Fig. 8). Concomitantly TCM significantly suppressed PMN respiratory
burst and phagocytosis (P ⬍ 0.05 vs. PMN incubated
with culture medium) (Fig. 9). These results indicate
that TCM-treated human PMN undergo an alteration
in phenotype resulting in an increased ability to bind to
tumor cells and to migrate through vascular endothelium, but a reduced cytocidal capacity. Furthermore, as
demonstrated in Fig. 10, both TCM-treated PMN and
migrated PMN had significantly delayed apoptotic
C819
C820
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
had no effect on TCM-treated, PMN-mediated tumor
cell migration.
The effect of human PMN treated with TCM, LPS, or
PMA on tumor cell transmigration, migrated tumor cell
viability, and proliferation. To examine the effect of
PMN activation on tumor cell transmigration, we stimulated human PMN with TCM; lipopolysaccharide
(LPS), which is known to be present in the circulation
following surgical insult; or phorbol 12-myristate 13acetate (PMA, a potent activator of PMN through PKC
activation); and compared PMN-assisted tumor cell
transmigration and migrated tumor cell viability and
proliferation. As shown in Table 1, LPS-activated PMN
or PMA-activated PMN also assisted tumor cell transendothelial migration, and the percentages of tumor
cell migration were similar to those mediated by TCMtreated PMN. There were no significant differences in
migrated tumor cell viability among PMN treated with
TCM, LPS, or PMA. However, migrated tumor cells
induced by LPS-treated PMN or PMA-treated PMN
had a significant reduction in cell proliferation (P ⬍
0.05 vs. TCM-treated PMN).
DISCUSSION
Tumor cell extravasation involves several sequential
events that include tumor cell arrest in the microvascular bed of target organs, adhesion to vascular endothelium lining the microvessels, transmigration
through the endothelial barrier, attachment to the
subendothelial basement membrane, invasion into extracellular matrix by local proteolysis, and finally metastatic colonization in the target organ. Adhesion of
tumor cells to vascular endothelium is a prerequisite
for tumor cell extravasation. Recent evidence suggests
that tumor cell attachment to vascular endothelium is,
at least in part, the result of adhesive interactions
between blood-borne tumor cells and cell surface pro-
Table 1. Comparison of PMN-assisted tumor cell
transmigration, migrated tumor cell viability and
proliferation among PMN treated with TCM,
LPS, or PMA
PMN ⫹ TCM
Tumor cell transmigration, %
Tumor cell viability, %
Tumor cell proliferation, %
of control
PMN ⫹ LPS
PMN ⫹ PMA
12 ⫾2.8
98 ⫾ 7.4
13 ⫾ 3.1
95 ⫾ 10
14 ⫾ 4.0
96 ⫾ 8.6
102 ⫾ 10
64 ⫾ 18*
55 ⫾ 21*
Data are expressed as the means ⫾ SD from 5 separate experiments. Isolated human polymorphonuclear neutrophils (PMN) were
incubated with tumor-conditioned medium (TCM), lipopolysaccharide (LPS, 100 ng/ml), or phorbol 12-myristate 13-acetate (PMA, 100
ng/ml) for 60 min. PMN-assisted MDA-MB-231 tumor cell transmigration, migrated tumor cell viability, and proliferation were assessed as described in MATERIALS AND METHODS. The statistical significance was compared with PMN ⫹ TCM, * P ⬍ 0.05.
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
Fig. 11. The effect of different blocking MAbs on TCM-treated PMNmediated tumor cell migration. Isolated human PMN and MDA-MB231 cells were pretreated with ICRF44 (10 ␮g/ml), TS1/18 (10 ␮g/ml),
P2A4 (20 ␮g/ml), and 4B4 (5.0 ␮g/ml) for 30 min, and TCM-treated
PMN-mediated MDA-MB-231 cell migration after 6 h of incubation
was assessed as described in MATERIALS AND METHODS. Data are
expressed as the means ⫾ SD from 4 separate experiments. The
statistical significance was compared with PMN pretreated with
culture medium, *P ⬍ 0.05.
teins expressed by vascular endothelium such as granule membrane protein 140 and its ligand (2), connexin43 proteins (5), vascular adhesion molecule-1
(VCAM-1) receptor ␣4␤1 and VCAM-1 (16), and selectins (18). Furthermore, several tumor cell lines have
been found to impair endothelial integrity by inducing
endothelial cell retraction (9) and apoptosis (11) after
contact between tumor cells and vascular endothelium,
which may facilitate tumor cell extravasation.
Tumor cell transmigration through the endothelial
barrier appears to be a critical step in the process of
tumor cell extravasation and subsequent metastatic
growth. However, the mechanisms involved in this
process remain unclear. The human breast adenocarcinoma cell line MDA-MB-231 cell line has been found
to be highly metastatic in vivo and highly invasive on
reconstituted basement membrane in vitro (3). These
tumor cells have also been found to cause endothelial
cell apoptosis, thus altering endothelial integrity (11).
In the present study, we investigated the hypothesis
that these cells may utilize proinflammatory leukocytes to facilitate transmigration. We found that MDAMB-231 cells failed to cross normal macro- and microvascular endothelial monolayers, suggesting that these
tumor cells are unable to complete the process of extravasation independently in this in vitro model. FACScan analysis demonstrated that MDA-MB-231 cells
expressed high levels of ICAM-1 and middle levels of
␤1-integrin, but they did not express ␤2-integrins,
namely, CD11a/CD18 and CD11b/CD18 and ␤3-integrin. As both CD11a/CD18 and CD11b/CD18 are necessarily required for the process of transendothelial
migration (6, 19), the lack of ␤2-integrin expression in
MDA-MB-231 cells may, at least in part, account for
the failure of these tumor cells to migrate through
endothelial monolayers.
The onset and progression of tumor metastasis is a
highly complicated and sequential multistep process.
Initiation of the metastatic process requires tumor
cells to undergo a sequence of extensive interactions
with a variety of host cells to achieve metastatic proliferation. Although it is generally agreed that the
host-tumor cell interaction is a defensive one, a num-
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
tumor cell migration. These results support our speculation that TCM-treated PMN-mediated MDA-MB231 cell transendothelial migration most likely occurs
through adhesive interactions between PMN and tumor cells. Furthermore, these cells possess proliferative potential following the migration process, and this
finding has potential clinical relevance.
The interesting finding in this study is that activation of human PMN by TCM, LPS, or PMA assists
tumor cell transendothelial migration; however, migrated tumor cells mediated by LPS-activated PMN or
PMA-activated PMN have an impaired ability to proliferate. Although the exact mechanisms involved are
unclear, the possible explanation for this phenomenon
may correlate with an increased cytocidal function in
PMN after LPS or PMA stimulation. Reactive oxygen
species released from LPS-activated PMN or PMAactivated PMN may affect tumor cell proliferation by
inhibition of cyclin A.
Together, our data support the concept that the
consequence of interactions between PMN and tumor
cells may support rather than inhibit tumor progression, since tumor cells can exploit PMN and alter their
function to increase their metastatic ability. Confirmation of this finding would have important clinical implications, especially in the perioperative period, when
tumor cells are known to be present in the circulation
and a transient leukocytosis occurs.
REFERENCES
1. Aeed PA, Nakajima M, and Welch DR. The role of polymorphonuclear leukocytes (PMN) on the growth and metastatic
potential of 13762NF mammary adenocarcinoma cells. Int J
Cancer 42: 748–759, 1988.
2. Aruffo A, Dietsch MT, Wan H, Hellstrom KE, and Hellstrom I. Granule membrane protein 140 (GMP 140) binds to
carcinomas and carcinoma-derived cell lines. Proc Natl Acad Sci
USA 89: 2292–2296, 1992.
3. Belloc C, Lu H, Fridman R, Legrand Y, and Menashi S. The
effect of platelets on invasiveness and protease production of
human mammary tumor cells. Int J Cancer 60: 413–417, 1995.
4. Crissman JD, Hatfield J, Schaldenbrand M, Sloane BF,
and Honn KV. Arrest and extravasation of B16 amelanotic
melanoma in murine lungs: a light and EM study. Lab Invest 53:
470–478, 1985.
5. El Sabban ME and Pauli BU. Adhesion-mediated gap junctional communication between lung-metastatic cancer cells and
endothelium. Invasion Metastasis 14: 164–176, 1994.
6. Godin C, Caprani A, Dufaux J, and Flaud P. Interaction
between neutrophils and endothelial cells. J Cell Sci 106: 441–
452, 1993.
7. Hangan D, Morris VL, Boeters L, von Ballestrem C, Uniyal
S, and Chan BM. An epitope on VLA-6 (alpha 6 beta 1) integrin
involved in migration but not adhesion is required for extravasation of murine melanoma B16F1 cells in liver. Cancer Res 57:
3812–3817, 1997.
8. Haslett C. Resolution of acute inflammation and the role of
apoptosis in the tissue fate of granulocytes. Clin Sci (Colch) 83:
639–648, 1992.
9. Honn KV, Tang DG, Grossi IM, Renaud C, Duniec ZM,
Johnson CR, and Diglio CA. Enhanced endothelial cell retraction mediated by 12(S)-HETE: a proposed mechanism for the role
of plates in tumor cell metastasis. Exp Cell Res 210: 1–9, 1994.
10. Jaffe EA, Nachman RL, Becker CG, and Minick CR. Culture of human endothelial cells derived from umbilical veins:
identification by morphologic and immunologic criteria. J Clin
Invest 52: 2745–2756, 1973.
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
ber of studies have shown that tumor cell-platelet (3,
19) and tumor cell-leukocyte (1, 14) interactions may in
fact support tumor cell invasion and the creation of the
optimal environment for tumor growth at the metastatic site. Using an in vitro transendothelial model,
we have demonstrated in this study that TCM-treated
human PMN assisted migration of MDA-MB-231 cells
across the endothelial barrier, indicating a supportive
role for PMN in tumor cell extravasation and metastasis. Because naive human PMN failed to aid tumor cell
transmigration, we wished to investigate the effect of
TCM on PMN. We found that TCM suppressed PMN
cytocidal function (respiratory burst and phagocytosis)
and concomitantly upregulated PMN adhesion receptor (CD11b/CD18) expression. This may indicate that
this metastatic breast cancer cell line secretes factors
capable of altering the phenotype of PMN to facilitate
transmigration while preventing PMN-directed tumor
cell damage.
Normal human PMN have a short half-life both in
vivo and in vitro consequent to the constitutive activation of an apoptotic cell death program (8). In this
study, TCM-treated PMN delayed apoptotic rates.
Again, one may speculate that this phenomenon is
beneficial to the metastatic process. Although it is
uncertain which of the MDA-MB-231 cell-producing
soluble factors present in TCM is responsible for altered human PMN function, tumor cell-produced granulocyte-macrophage colony-stimulating factor and/or
IL-3 have been found, in previous reports, to be responsible for potentiating tumor metastases induced by
tumor-elicited PMN (12).
There are two possible ways that PMN may assist
tumor cell migration across an endothelial barrier. The
first possibility is that PMN causes endothelial injury
by releasing proteases and reactive oxygen species to
increase microvascular permeability for tumor cell extravasation, as seen in a tumor-bearing murine model
in which PMN-mediated pulmonary vascular injury
facilitated tumor cell metastasis in the lung (15). Our
results from an endothelial monolayer permeability
experiment indicate that the increased MDA-MB-231
cell transmigration mediated by TCM-treated human
PMN was not due to the impaired integrity of the
endothelial monolayer. The other possibility is an interaction that may occur between PMN and tumor cells
in which PMN, through an adhesion receptor-dependent mechanism, may bind to tumor cells and facilitate
their migration through vascular endothelium. In the
present study, the evidence that MDA-MB-231 cells
expressed high levels of ICAM-1 and that TCM upregulated PMN CD11b and CD18 expression indicates that
TCM-treated PMN possess the appropriate phenotype
to bind and transport tumor cells. The finding that IL-8
increased naive PMN, but not tumor cell migration,
suggests that the binding of PMN to tumor cells is a
prerequisite for PMN-assisted tumor cell migration.
Furthermore, by using different blocking MAbs, we
have found that CD11b and CD18 expressed on human
PMN and ICAM-1 expressed on MDA-MB-231 cells
were involved, at least in part, in PMN-facilitated
C821
C822
NEUTROPHILS AND TUMOR CELL EXTRAVASATION
18. Saiki I, Koike C, Obata A, Fugii H, Murata J, Kiso M,
Hasegawa A, Komazawa H, Tsukada H, Azuma I, Okada S,
and Oku N. Functional role of sialyl Lewis X and fibronectinderived RGDS peptide analog on tumor cell arrest in lungs
followed by extravasation. Int J Cancer 65: 833–839, 1996.
19. Smith CW and Anderson DC. PMN adhesion and extravasation as a paradigm for tumor cell dissemination. Cancer Metastasis Rev 10: 61–78, 1991.
20. Smith WB, Gamble JR, Clark-Lewis I, and Vadas MA.
Interleukin 8 induces neutrophil transendothelial migration.
Immunology 72: 65–72, 1991.
21. Starkey JR, Liggitt HD, Jones W, and Hosick HL. Influence
of migratory blood cells on the attachment of tumor cells to
vascular endothelium. Int J Cancer 34: 535–543, 1984.
22. Van Netten JP, Ashmead BJ, Parker RL, Thornton IG,
Fletcher C, Cavers D, Coy P, and Brigden ML. Macrophagetumor cell association: a factor in metastasis of breast cancer?
J Leukoc Biol 54: 360–362, 1993.
23. Wang JH, Sexton DM, Redmond HP, Watson RWG, Croke
DT, and Bouchier-Hayes D. Intercellular adhesion molecule-1
(ICAM-1) is expressed on human neutrophils and is essential for
neutrophil adherence and aggregation. Shock 8: 357–361, 1997.
24. Welch DR, Schissel DJ, Howrey RP, and Aeed PA. Tumorelicited polymorphonuclear cells, in contrast to “normal” circulating polymorphonuclear cells, stimulate invasive and metastatic potentials of rat mammary adenocarcinoma cells. Proc
Natl Acad Sci USA 86: 5859–5863, 1989.
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.2 on June 17, 2017
11. Kebers F, Lewalle JM, Desreux J, Munaut C, Devy L,
Foidart JM, and Noel A. Induction of endothelial cell apoptosis
by solid tumor cells. Exp Cell Res 240: 197–205, 1998.
12. McGary CT, Miele ME, and Welch DR. Highly metastatic
13762NF rat mammary adenocarcinoma cell clones stimulate
bone marrow by secretion of granulocyte-macrophage colonystimulating factor/interleukin-1 activity. Am J Pathol 147:
1668–1681, 1995.
13. Nicoletti I, Migliorati G, Pagliacci MC, Grignani F, and
Riccardi C. A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry.
J Immunol Methods 139: 271–279, 1991.
14. Orr FW and Mokashi S. Effect of leukocyte activation on the
formation of heterotypic tumor cell aggregates in vitro. Int J
Cancer 35: 101–106, 1985.
15. Orr FW and Warner DJA. Effects of systemic complement
activation and neutrophil-mediated pulmonary injury on the
retention and metastasis of circulating cancer cells in mouse
lungs. Lab Invest 62: 331–338, 1990.
16. Paavonen T, Tiisala S, Majuri ML, Bohling T, and Renkonen R. In vivo evidence of the role of alpha 4 beta 1-VCAM-1
interaction in sarcoma, but not in carcinoma extravasation. Int J
Cancer 58: 298–302, 1994.
17. Rosales C and Juliano RL. Signal transduction by cell adhesion receptors in leukocytes. J Leukoc Biol 57: 189–198, 1995.