Thrombin receptor expression and responsiveness of human

Thrombin receptor expression and responsiveness of human

JOURNAL OF CELLULAR PHYSIOLOGY 177:76–84 (1998)
Thrombin Receptor Expression and
Responsiveness of Human Monocytic Cells
to Thrombin Is Linked to Interferon-Induced
Cellular Differentiation
ANTONELLA NALDINI,1* LAURIE SOWER,2 VELIO BOCCI,1 BECKY MEYERS,2
2
AND DARRELL H. CARNEY
1
University of Siena, Institute of General Physiology, Siena, Italy
2
Department of Human Biological Chemistry and Genetics,
University of Texas Medical Branch, Galveston, Texas
Human thrombin has been shown to stimulate monocyte chemotaxis, phagocytosis, and interleukin (IL8) production, but the mechanisms responsible for stimulation
are not well defined. In some cells, thrombin stimulation of proliferation appears
to require both cleavage of the proteolytically activated receptor for thrombin
(PAR1) and activation of a nonproteolytically activated thrombin receptor (N-PAR),
while in others activation of either receptor alone may be sufficient for stimulation.
We, therefore, have initiated studies to address thrombin receptor expression and
cell responsiveness to thrombin in interferon gamma (IFNg)-differentiated and nondifferentiated U937 monocytic cells. Northern blot analysis shows that PAR1 expression is upregulated upon differentiation. Experiments with biotinylated and
125
I-thrombin show that specific thrombin binding is dramatically increased by
differentiation although it is not clear if this binding is to PAR1 or to a separate
binding component such as N-PAR which is present on fibroblasts and other cells.
Addition of thrombin at concentrations of 1–10 mg/ml (30–300 nM, concentrations
where specific thrombin binding is observed) stimulates proliferation of IFNg-differentiated U937 cells but not of undifferentiated U937 cells. Thrombin also stimulates
interleukin-6 (IL6) production in IFNg-differentiated U937 cells. Moreover, thrombin induces high levels of IL6, interleukin-1b (IL1b), and tumor necrosis factor-a
(TNFa) production by peripheral blood mononuclear cells (PBMC) and monocytes.
These results show that differentiated U937 cells and mature PBMC are responsive
to thrombin whereas nondifferentiated U937 are not. Further, this responsiveness
appears to correlate with expression of PAR1 and to a dramatic increase in specific
thrombin binding. That thrombin stimulates cytokine production and proliferation
in populations of differentiated monocytes suggests that thrombin may be an important regulator of inflammation and wound healing. J. Cell. Physiol. 177:76–
84, 1998. q 1998 Wiley-Liss, Inc.
In addition to its pivotal role in coagulation, thrombin interacts with many types of cells to induce a variety of cellular functions. Thrombin stimulates proliferation of fibroblasts (Carney and Cunningham, 1978),
endothelial cells (Gospodarowicz et al., 1978), neuronal
cells (Gurwitz and Cunningham, 1988), vascular
smooth cells (Weiss and Nuccitelli, 1992), and bone
marrow cells in the presence of colony stimulating factor-1 (CSF-1; Clohisy et al., 1990). Further, thrombin
has potential to affect local inflammatory response by
acting as a chemotactic agent for polymorphonuclear
cells (PMNs) and monocytes (Bar-Shavit et al., 1983;
Bizios et al., 1986) by stimulating the production of
monocyte chemotactic protein-1 (MCP-1; Colotta et al.,
1994), by stimulating proliferation of monocytes (Clohisy et al., 1990), by stimulating the release of interleukin-1 (IL1) and interleukin-8 (IL8; Jones and Geczy,
1990; Sower et al., 1996), and by enhancing monocyte
phagocytic activity (Sower et al., 1996). Thrombin also
may affect other aspects of inflammation and immunologic surveillance by stimulating T-cell activation and
cytokine production (Naldini et al., 1993) and modulation of both natural killer (NK) and lymphokine-activated killer (LAK) cell cytotoxicity (Naldini and Carney, 1996).
Many of these cellular events induced by thrombin
have been shown to require both proteolytic activation
Contract grant sponsor: M.U.R.S.T.; Contract grant sponsor: Consorzio Siena Ricerche Fund; Contract grant sponsor: NIH; Contract grant number: GM-47572.
*Correspondence to: Dr. Antonella Naldini, Institute of General
Physiology, University of Siena, Via Laterina 8, 53100 Siena,
Italy. E-mail: [email protected]
Received 5 December 1997; Accepted 20 February 1998
q 1998 WILEY-LISS, INC.
/ 893f$$0284
07-20-98 20:24:53
wlcpal
W Liss: JCP
JCP-10284
77
THROMBIN-INDUCED MONOCYTE ACTIVATION
of the G-protein-linked proteolytically activated receptor for thrombin (PAR1) and binding to or activation
of a separate nonproteolytically activated receptor (NPAR) (Carney et al., 1986; Grand et al., 1996). It has
been reported that the majority of NK and T cells and
monocytes express PAR1 (Kudahl et al., 1991; Tordai
et al., 1993). In certain fibroblasts and other cells, however, activation of PAR1 is not sufficient to stimulate
cell proliferation (Bar-Shavit et al., 1986; Brass et al.,
1991; Vouret-Craviari et al., 1992). Interestingly, certain cells including neutrophils, bind thrombin with
moderate/ high affinity (Kd Å Ç2 nM), yet they do not
express PAR1 (Bizios et al., 1986). In these neutrophils,
thrombin stimulates neutrophil chemotaxis by a mechanism that appears to be independent of proteolytic
activity or activation of PAR1 (Jenkins et al. 1995). We,
therefore, wanted to determine whether expression of
PAR1, N-PAR, or both correlated with thrombin stimulation of monocyte proliferation and cytokine production. For these studies, we compared PAR1 expression,
thrombin binding, and cell responsiveness in IFNg-differentiated and undifferentiated human monocytic
U937 cells derived from a histiocytic lymphoma (Sundström and Nilsson, 1976). It has been shown that IFNg,
which is an important macrophage activating factor,
induces changes in the U937 cell line that reflect cellular differentiation (Ucla et al., 1990; Roberts et al.,
1991). Therefore, U937 cells provide a convenient
model system to assess thrombin responsiveness during monocyte differentiation. These studies show that
thrombin induces proliferation and cytokine production
by IFNg-differentiated U937 cells, but not in undifferentiated U937. Interestingly, differentiation of these
cells by IFNg appears to upregulate the expression of
PAR1 and specific binding of thrombin to these cells.
Thus, thrombin may play an important role in regulating differentiated monocytic functions in these cells at
sites of tissue repair, inflammation, or even in atherosclerotic plaques.
MATERIALS AND METHODS
Reagents
Highly purified human a-thrombin (99% a form) was
a kind gift from Dr. J.W. Fenton (Albany, NY) (specific
activity 3,683 NIH U/mg protein). Another preparation
of highly purified human a-thrombin (specific activity
4,000 NIH U/mg protein) and 3-(4,5- dimethylthiazol2yl)-2,5-diphenyl tetrazolium bromide (MTT) were purchased from Sigma (St. Louis, MO). Recombinat human
IFNg was obtained from Genentech (San Francisco, CA)
with a specific activity of 2 1 107 IU/mg protein.
Cell separation
Peripheral blood from healthy donors was used as a
source of mononuclear cells (PBMC) and was applied
directly to a gradient of Ficoll-Hypaque (Pharmacia,
Piscataway, NJ) as previously described (Naldini et al.,
1997). Monocyte-depleted PBMCs were obtained as follows. PBMCs were passed over nylon wool columns and
the eluted cells were further incubated once or twice
on plastic in culture grade flasks (Costar, Cambridge,
MA) at 377C for 2 hr in RPMI (GIBCO, Grand Island,
NY) and 2% heat-inactivated fetal calf serum (FCS).
Nonadherent cells were collected, washed three times
with HYQ-CCM1 serum-free medium (Hyclone, Logan,
UT), and resuspended for culture (2 1 105/100 ml). The
/ 893f$$0284
07-20-98 20:24:53
composition of cells in the suspension was Ç90% T lymphocytes (CD3/), Ç10% NK cells (CD16/), 0% monocytes (CD14/), and 0% B lymphocytes (CD19/) as assessed by flow cytometry. Monocyte-enriched populations were obtained by plating PBMC (5 1 105 cells
per well) in 200 ml medium in 96-well tissue culture
microplates and allowed to adhere for 2 hr at 377C.
Nonadherent cells were removed by washing eight
times with phosphate-buffered saline (PBS). Fresh serum-free medium (100 ml/well) was added and cells
were cultured for the indicated intervals as described
below. Neutral red phagocytosis assays revealed that
95% of the adherent cells were monocytes (Gardner and
Remington, 1978).
U937 cell culture and IFNg treatment
U937 cells were cultured in RPMI-1640 supplemented
with 10% heat-inactivated FCS, 2 mM glutamine, 100
IU/ml penicillin, and 100 mg/ml streptomycin. Cells were
induced to differentiate by supplementing the medium
with 200 IU/ml IFNg for 10 days as previously described
(Roberts et al., 1991). Every 2 days the cells were resuspended in fresh medium containing IFNg to maintain
the cell concentration between 1–5 1 105 cells/ml. After
10 days, cells were washed extensively and cultured as
described below. To determine whether IFNg treatment
induced differentiation in U937 cells, cell proliferation
was routinely tested, showing a decrease of about 50%
than controls as previously described (Oberg et al., 1991).
Moreover, U937 cells were analyzed by a Nitro Blue Tetrazolium (NBT) test in the presence or absence of 1.67
mM of 12-O- tetradecanoylphorbol-13-acetate (TPA) as
previously reported (Roberts et al., 1991). The percentage
of cells containing formazan in each sample was determined by counting at least 300 cells (undifferentiated
U937 cells, Ç1%; IFNg-differentiated U937 cells, Ç50%).
Northern blot analysis
Total RNA was extracted from undifferentiated and
IFNg-differentiated U937 cells using TRI REAGENT
(Molecular Research Center, Inc., Cinncinnati, OH),
separated on a 5.4% formaldehyde, 1% agarose gel, and
blotted onto nylon membrane (Micron Separations Inc.,
Westboro, MA). The membrane was UV crosslinked.
Human a-thrombin receptor probe (PAR1) (a kind gift
from Dr. M.S. Runge, Galveston, TX) or G3PDH obtained from Clontech (Palo Alto, CA) was then radiolabeled with 32P-dCTP obtained from Amersham (Arlington Heights, IL) using a Decaprime labeling kit (Ambion Inc., Austin, TX). After hybridization in 50%
formamide, 51 SSPE, 0.1% SDS, 51 Denhardt’s, and
100 mg/ml sheared salmon sperm DNA at 557C overnight (18 hr), the nylon membrane was washed under
high stringency conditions (0.11 SSC, 1% SDS) and
exposed to a phosphoimaging screen overnight (18 hr;
Molecular Dynamics, Sunnyvale, CA). Bands were
quantitated using the Image Quant software available
within the phosphoimager.
Magnetic bead sorting
1 1 106 undifferentiated and IFNg-differentiated
U937 cells were placed in serum-free media for 2 hr.
Cells were then centrifuged for 9 min at 1,000 rpm
and resuspended in Dulbecco modified Eagle medium:Ham’s F12 nutrient mixture (1:1) (DV) with 15
mM Hepes, 1% bovine serum albumin (BSA) at 1 1
wlcpal
W Liss: JCP
JCP-10284
78
NALDINI ET AL.
TABLE 1. Effect of thrombin and FCS on undifferentiated
U937 cells
Medium
Thrombin
FCS
1
114 { 1.5
130 { 4.9*
100 { 5.3
1
Results are expressed as O.D. %. 105 U937 cells were cultured in 100 ml of
serum-free medium (HYQ-CMM1) with the addition of either thrombin (10 mg/ml)
of FCS (0.25%) for 1 hr at 377C. An additional 100 ml of medium supplemented
with FCS (1%) was added. Following 72 hr, 100 ml of culture medium was removed from each well and 10 ml of MTT was added. After 4 hr incubation, 100
ml of acid propan-2-ol was added and the O.D. was assessed as described in
Materials and Methods. Data represent a typical experiment performed in quadruplicate.
*Statistically significant (P õ 0.05) difference between O.D.% of cultures treated
with either thrombin or FCS vs. untreated cultures.
Filters were counted in a gamma counter and specific
binding was determined by subtracting the nonspecific
binding from total counts.
Fig. 1. Thrombin enhances IFNg-differentiated U937 proliferation.
105 U937 and IFNg- differentiated U937 cells were cultured in 100 ml
of serum-free medium (HYQ-CMM1) without (open bars) or with 5
mg/ml of thrombin (filled bars) for 1 hr at 377C. An additional 100 ml
of medium supplemented with FCS (1%) was added. Following 72 hr,
100 ml of culture medium was removed from each well and 10 ml of
MTT was added. After 4 hr incubation, 100 ml of acid propan-2-ol
was added and the O.D. was assessed as described in Materials and
Methods. Data represent the M { SEM of three independent experiments performed in quadruplicate. Asterisk indicates statistically significant (P õ 0.05) differences between O.D.% of cultures treated with
thrombin vs. untreated cultures.
107/ml. 1 1106 cells were aliquoted into 1.5 ml microfuge tubes and incubated at 227C for 10 min. 60 ng/ml
biotinylated a-thrombin was added and the incubation
continued for 90 min at 227C. Streptavidin-coated magnetic beads (CPG, Lincoln Park, NJ) were washed and
resuspended in binding buffer (10 mM phosphate, pH
7.5) at 0.56 mg/ml. Beads were added such that the
bead:cell ratio was approximately 3:1 (0.225 mg) and
incubation was continued for 1 hr with rotation at 227C.
Cells were magnetically separated and the supernatant
was aspirated. Magnetic beads with or without cells
attached were washed two to three times by adding 1
ml of washing buffer (10 mM phosphate, pH 7.5, 1 M
NaCl, 0.1% BSA) followed by magnetic separation.
After the final wash, beads were resuspended in 1 ml of
washing buffer diluted in 9 ml PBS-azide and counted.
More than 99% of the cells were viable following this
assay as measured by trypan blue exclusion.
125
I-thrombin binding
6
1 1 10 (116 mg of protein) U937 or IFNg-differentiated U937 (110 mg of protein) were placed in serumfree medium for 2 hr and then resuspended in 0.1 ml
DV-Hepes/1% BSA in 1.5 ml microfuge tubes and incubated for 10 min at 227C. 125I-thrombin (60 ng/ml) was
added to each tube to determine total binding. Nonspecific binding was assessed by adding 100-fold excess of
cold a-thrombin to 60 ng/ml I125 -thrombin. Cells were
incubated for 2 hr at 227C and then filtered through
Whatman GF/C filters followed by washing with PBS.
/ 893f$$0284
07-20-98 20:24:53
Proliferation assays
The proliferative responses of U937 cells to thrombin
were performed using the MTT colorimetric method
(Mosmann, 1983). Briefly, U937 or IFNg-differentiated
U937 cells (105 cells/well) were incubated in 100 ml serum-free medium, in 96-well microtiter plates at 377C
in the presence of varying concentrations of thrombin.
One hour later, 100 ml of medium supplemented with
2% FCS was added and incubation continued. Following indicated time points, 100 ml of culture medium
was removed and cells were treated with 10 ml of MTT
solution. Four hours later, 100 ml of acid propan-2-ol
(0.04 M HCl in propan-2-ol) was added to dissolve the
formazan product. The microplates were read using a
microelisa reader at 570 nm using a reference wavelength of 630 nm and a calibration setting of 1.99.
Cytokine assays
IL6, IL1b, TNFa, and IL8 were assessed from cellfree supernatants by enzyme-linked immunosorbent
assay (ELISA). Briefly, U937 or IFNg-differentiated
U937 cells, PBMC, monocytes, and monocyte-depleted
PBMC were cultured with and without thrombin at
the indicated concentrations in 100 ml of serum-free
medium as described above. One hour later, 100 ml of
medium containing 2% FCS was added and incubation
continued. At the appropriate intervals, cell culture supernatants were harvested and placed at -207C pending
assay. ELISA kits were obtained from Biosource International (Camarillo, CA) and R&D Systems (Minneapolis, MN) and did not show cross-reactivity with other
cytokines (õ0.0005%). Minimum detectable doses were
as follows: IL6 Å 2 pg/ml; IL1b Å 3 pg/ml; TNFa Å 4.4
pg/ml; IL8 Å 3 pg/ml. Coefficients of variation for intraassay precision were õ5%.
Statistics
Unless otherwise stated, results are expressed as
percent of control. Data represent the mean { SEM.
Statistical analysis was by Student’s two-tailed t-test.
RESULTS
Effects of thrombin on differentiated
and undifferentiated U937 cells
To evaluate whether thrombin differentially affects
the proliferation of monocyte-differentiated and undifferentiated cells, IFNg-treated and control U937 cells
wlcpal
W Liss: JCP
JCP-10284
79
THROMBIN-INDUCED MONOCYTE ACTIVATION
Fig. 2. A: Dose response of thrombin enhancement of IFNg-differentiated U937 proliferation. IFNg-differentiated U937 cells were incubated with HYQ-CMM1 medium or varying concentrations of thrombin for 1 hr at 377C. An additional 100 ml of medium supplemented
with FCS (1%) was added. Following 72 hr, 100 ml of culture medium
was removed from each well and 10 ml of MTT was added. After 4 hr
incubation, 100 ml of acid propan-2-ol was added and the O.D. was
assessed as described in Materials and Methods. Data are the M {
SEM of three independent experiments performed in quadruplicate.
Asterisks indicate statistically significant (P õ 0.05) differences between O.D.% of cultures treated with thrombin vs. untreated cultures.
B: Kinetics of thrombin-induced proliferation. IFNg-differentiated
U937 cells were incubated with HYQ-CCM1 medium or thrombin (10
mg/ml) for 1 hr at 377C, then supplemented with FCS (1%). After 24,
48, and 72 hr, 100 ml of culture medium was removed from each well
and 10 ml of MTT was added. After 4 hr incubation, 100 ml of acid
propan-2-ol was added and the O.D. was assessed as described under
Materials and Methods. Data are the M { SEM of three independent
experiments performed in quadruplicate. Asterisks indicate statistically significant (P õ 0.05) differences between O.D.% of cultures
treated with thrombin vs. untreated.
were incubated in the presence or absence of thrombin
and proliferation was assessed using the MTT colorimetric assay. Treatment of IFNg-differentiated U937
cells with thrombin resulted in a significant increase
in proliferative response (Fig. 1). The increase in proliferation induced by thrombin was approximately 160%
greater than those cells treated with medium alone. It
should be noted that in these experiments, there was
a small increase in the proliferative response of undifferentiated U937, even in the presence of 10 mg/ml
thrombin; however FCS was able to increase significantly the cell proliferation as shown in Table 1.
Thrombin induction of proliferation was concentration
dependent with optimal proliferation observed using
5 – 10 mg/ml of thrombin (Fig. 2A). This concentration
of thrombin has also been shown to be necessary for Tlymphocyte and monocyte thrombin-induced stimulatory effects (Clohisy et al., 1990; Naldini et al., 1993).
The kinetics of the thrombin-induced proliferation of
IFNg-treated U937 cells are shown in Figure 2B.
Thrombin treatment induced significant proliferation
of IFNg-differentiated U937 cells first observed at 24 hr
which continued to increase to 72 hr of culture. These
results show that thrombin stimulates proliferation of
IFNg-differentiated U937 cells in a time and dose-dependent manner, but that thrombin does not stimulate
proliferation in undifferentiated U937 cells. Moreover,
the effect of thrombin of differentiated U937 cells still
persists when IFNg is left in the culture (Table 2).
/ 893f$$0284
07-20-98 20:24:53
Thrombin enhances IL6 production
by U937 cells
To further investigate thrombin enhancement of
monocyte activation, we investigated whether thrombin treatment of IFNg-differentiated U937 cells resulted in enhanced IL6 production. IFNg-differentiated
U937 were treated with or without thrombin and assessed for IL6 production. Thrombin induced IL6 pro-
wlcpal
W Liss: JCP
JCP-10284
80
NALDINI ET AL.
TABLE 2. Effect of thrombin on differentiated U937 cells in the
presence of IFNg
TABLE 3. Thrombin enhances IL6 production in human
monocytes1
Thrombin (mg/ml)
Medium
1001 { 3.5
1
2.5
111 { 1.9
135 { 2.5*
1
Results are expressed as O.D.%. 105 differentiated U937 cells were cultured in
100 ml of serum-free medium (HYQ-CMM1) with addition of thrombin, in the
presence of IFNg (100 IU/ml) for 1 hr at 377C. An additional 100 ml of medium
supplemented with FCS (1%) was added. Following 72 hr, 100 ml of culture
medium was removed from each well and 10 ml of MTT was added. After 4 hr
incubation, 100 ml of acid propan-2-ol was added and the O.D. was assessed
as described in Materials and Methods. Data represent a typical experiment
performed in quadruplicate.
*Statistically significant (P õ 0.05) difference between O.D.% of cultures treated
with thrombin vs. untreated cultures.
duction in a dose-dependent manner (Fig. 3A). Although thrombin was ineffective at a concentration of
1 mg/ml, 10 mg/ml of thrombin induced a twofold increase in the IL6 production by IFNg-differentiated
U937 cells. These data are in agreement with our proliferation experiments (Figs. 1, 2) as well as with previous
studies using T lymphocytes and PBMC (Naldini et al.,
1993). The kinetics of IL6 production are presented in
Figure 3B. Thrombin-induced IL6 production was detected as early as 24 hr posttreatment (80% increase)
and increased slightly at 48 hr of culture (100% increase) with a decline at 72 hr (60% increase). In these
experiments, thrombin treatment of undifferentiated
U937 did not result in an increase of IL6 production
(data not shown).
Thrombin enhances cytokine production
by peripheral blood monocytes
Since thrombin induced IL6 production in IFNg-differentiated U937 cells, we next investigated thrombin’s
effect on cytokine production in human peripheral
blood monocytes. Monocytes were treated with thrombin and cell-free supernatants from these cultures were
assessed for IL6, IL1b, TNFa, and IL8 production. IL6
production was dose dependent following thrombin
treatment of monocytes (Table 3). These results are
consistent with the results obtained in IFNg-differentiated U937 (Fig. 3). The increase in IL6 production induced by thrombin (10 mg/ml) was more than 60%
greater than control cultures without thrombin treatment. To further confirm the role of monocytes in the
thrombin-induced IL6 production, we performed experiments using unfractionated PBMC and monocyte-depleted PBMC (Table 4). Thrombin induced a threefold
increase in IL6 production by unfractionated PBMC
(more than a 120% increase) when compared with
thrombin-treated monocyte-depleted PBMC which produced no IL6. It should be noted that in these experiments monocyte-depleted PBMC contain Ç90% T lymphocytes (CD3/), 10% NK cells (CD 16/), and less than
0.1% monocytes (CD14/). Thrombin induction of cytokine production by monocytes was not limited to IL6.
Monocytes treated with thrombin produced twofold
more IL1b and TNFa when compared to cultures
treated with medium alone (Fig. 4). Interestingly, in
monocytes just as in IFNg-differentiated U937 (data
not shown), thrombin did not induce IL8 production.
Thus, thrombin- induced cytokine production in human
monocytes is comparable to that seen in U937 cells.
These results suggest a direct role for thrombin in regu-
/ 893f$$0284
07-20-98 20:24:53
Culture
conditions
Thrombin (mg/ml)
0
1
5
10
IL6 production
IL6 (pg/ml)
% of control
1,512
1,702
1,827
2,124
{
{
{
{
126
96
62*
109*
100
134
143
167
{
{
{
{
3
17
17*
22*
1
Human monocytes were incubated with serum-free medium (HYQ-CCM1) with
or without varying concentrations of thrombin for 1 hr at 377C, then supplemented with the same medium / FCS (1%). After 48 hr, cell-free supernatants
were obtained and pooled from triplicate cultures. Results are expressed as
pg/ml of IL6 and as percent of control. Data presented are the M { SEM of four
independent experiments.
*Statistically significant (P õ 0.05) differences between IL6 produced in cultures
treated with thrombin vs. untreated parallel cultures.
lating the production of inflammatory cytokines by human monocytes.
Correlation between increased response to
thrombin and thrombin receptor expression
A number of receptors appear to be expressed on
various cells to which thrombin may bind to initiate
proliferative and cytokine releasing signals. We therefore set out to determine if IFNg- induced differentiation caused upregulation of either PAR1 or the specific
binding of thrombin to cells. Monocytes have been
shown to express PAR1 (Hoffman and Church, 1993).
To determine whether PAR1 is expressed and upregulated during monocyte differentiation, U937 cells were
treated with and without IFNg for 10 days. Total RNA
was purified and subjected to Northern blot analysis
as described in the Materials and Methods section.
U937 cells differentiated into monocytes expressed
three to fivefold more PAR1 message than undifferentiated U937 cells (Fig. 5). This increase is not seen in
G3PDH. Although cells treated with IFNg express a
spliced G3PDH message, quantitation shows that the
hybridization to the PAR1 message is two to threefold
higher than that for either of the G3PDH bands. Thus,
these results indicate that upon differentiation, PAR1
expression is increased.
Because there is evidence for moderate/high-affinity
receptor binding for thrombin on many cells (Carney
et al., 1992b), including those that do not express PAR1,
we wanted to determine whether specific thrombin
binding also increases with monocyte differentiation.
We have recently developed a nonradioactive assay
that measures retention of thrombin on the surface of
cells. The results of our nonradioactive assay are shown
in Figure 6A. Undifferentiated and IFNg-differentiated
U937 cells were incubated with or without biotinylated
thrombin. Those cells that bound biotinylated thrombin
were separated from those that did not bind thrombin
by using streptavidin-coated magnetic beads. As shown
in Figure 6A, IFNg-differentiated cells had a threefold
increase in the number of cells binding thrombin when
compared with the number of undifferentiated cells
binding thrombin. To determine whether the binding
of thrombin to differentiated U937 cells was a specific
interaction, we initiated 125I-thrombin binding studies.
Figure 6B shows the increase in specific binding of 125Ithrombin to IFNg-differentiated vs. undifferentiated
U937 cells. As shown with this assay as well, there is a
wlcpal
W Liss: JCP
JCP-10284
81
THROMBIN-INDUCED MONOCYTE ACTIVATION
Fig. 3. A: Dose response of thrombin enhancement of IFNg-differentiated U937 IL6 production. IFNg-differentiated U937 cells were incubated with HYQ-CMM1 medium or varying concentrations of thrombin for 1 hr at 377C, then supplemented with FCS (1%). After 48
hr, cell-free supernatants were obtained and pooled from triplicate
cultures. Data presented are expressed as % of control and are the
M { SEM of three independent experiments. Asterisks indicate statistically significant (P õ 0.05) differences between IL6 produced in
cultures treated with thrombin vs. untreated parallel cultures. B:
Kinetics of thrombin-induced IL6 production. IFNg-differentiated
U937 cells were incubated with HYQ-CMM1 medium or varying concentrations of thrombin for 1 hr at 377C, then supplemented with FCS
(1%). After 24, 48, and 72 hr, cell-free supernatants were obtained
and pooled from triplicate cultures. Data presented are expressed as
% of control and are the M { SEM of three independent experiments.
Asterisks indicate statistically significant (P õ 0.05) differences between IL6 produced in cultures treated with thrombin vs. untreated
cultures.
TABLE 4. Effects of thrombin on IL6 production in PBMC1
IL6 production
Culture conditions
Medium
Thrombin
Unfractionated PBMC
IL6 (pg/ml)
% of control
1,022 { 785
3,210 { 1,361
100 { 21
223 { 44*
Monocyte-depleted PBMC
IL6 (pg/ml)
% of control
õ2
õ2
n.a.2
n.a.
1
Unfractionated and monocyte-depleted PBMC from the same buffy coat were cultured with serum-free medium (HYQ-CMM1) with and without 10 mg/ml of
thrombin for 1 hr at 377C, then supplemented with the same medium / FCS (1%). After 48 hr, cell-free supernatants were obtained and pooled from triplicate
cultures. Results are expressed as pg/ml of IL6 and as percent of control. Data presented are the M { SEM of three independent experiments.
2
Not applicable.
*Statistically significant (P õ 0.05) differences between IL6 produced in cultures treated with thrombin vs. untreated parallel cultures.
threefold increase in specific thrombin binding to IFNgdifferentiated U937 cells. Thus, in both nonradioactive
thrombin binding assays and assays with 125I-thrombin, the thrombin binding to U937 cells is shown to
increase by approximately threefold when these cells
are differentiated toward monocytes.
/ 893f$$0284
07-20-98 20:24:53
DISCUSSION
Thrombin has been shown to mediate a variety of
biological effects (Shuman, 1986). Many of the functions mediated by thrombin have also been shown to
play important roles in wound healing and in inflam-
wlcpal
W Liss: JCP
JCP-10284
82
NALDINI ET AL.
Fig. 5. Upregulation of PAR1 expression in IFNg-differentiated U937
cells. Total RNA from U937 (lane A) and IFNg-differentiated U937
cells (lane B) were treated as described in Materials and Methods.
Quantitation was achieved by using the Image Quant program of the
phosphoimager. Percent area for the top blot for lane A: 5; lane B: 38.
Percent area for the bottom blot for lane A: 25; lane B, upper band:
15; lower band: 18.
Fig. 4. Effects of thrombin on cytokine production by human monocytes. Human monocytes were incubated with HYQ-CMM1 medium
(open bars) or 5 mg/ml of thrombin (filled bars) for 1 hr at 377C, then
supplemented with FCS (1%). After 48 hr, cell-free supernatants were
obtained and pooled from triplicate cultures. Data presented are expressed as % of control and are the M { SEM of three independent
experiments. Asterisks indicate statistically significant (P õ 0.05)
differences between cytokines produced in cultures treated with
thrombin vs. untreated cultures.
mation (Carney et al., 1992a). In this report, we now
show that with differentiation of U937 cells toward
monocytes, cells become competent to respond to
thrombin with cytokine production and enhanced cell
proliferation. With this differentiation, thrombin binding and PAR1 expression both increase by approximately threefold. Thus, as monocytes differentiate, it
appears that they become more responsive to thrombin.
Clohisy et al. (1990) have shown that thrombin was
not mitogenic for normal murine bone marrow-derived
macrophages. However, thrombin did greatly enhance
macrophage proliferation in response to CSF-1. These
results are similar to our findings using human
PBMCs, monocytes, and U937 cells. Thrombin was
rarely mitogenic to undifferentiated U937 cells but did
greatly enhance proliferation of IFNg-differentiated
U937. In agreement with our findings, Bar-Shavit et
al. (1987) reported that thrombin was chemotactic for
the human leukemic cell line HL60 which was induced
to differentiate along the monocytic lineage.
Thrombin is capable of inducing various cells to produce a number of proinflammatory cytokines which mediate induction of different biological responses. Endothelial cells stimulated with thrombin synthesize platelet-derived growth factor (PDGF; Harlan et al., 1986),
IL1 (Stern et al., 1985), prostaglandins (Weksler et al.,
1978), and plasminogen activators (Levin et al., 1984).
Thrombin was also shown to greatly enhance IL1 production by lipopolysaccharide (LPS)-stimulated macrophages (Jones and Geczy, 1990). We recently reported
that thrombin can activate human T lymphocytes and
enhance cytokine (IL6 and IL2) production by PBMC
(Naldini et al., 1993). IL6 is a pleiotropic cytokine that
/ 893f$$0284
07-20-98 20:24:53
can be produced by a variety of cells and acts on a
variety of tissues, exerting growth-inducing, growthinhibitory, and differentiation-inducing effects, depending on the target cell (Van Snick, 1990). In our
studies, thrombin did not induce IL6 production by undifferentiated U937 or monocyte-depleted PBMC, but
did induce IL6 by IFNg-differentiated U937 cells. These
results were unexpected as it has been shown that
monocytes do not produce IL6 upon thrombin stimulation (Sower et al., 1996); however, the time points when
supernatants were collected differ between the two
studies. In these experiments, supernatants were collected at 48 hr, whereas in the previous studies, supernatants were collected at 8 hr. While the production of
IL6 is apparent, these results suggest that the cytokine
production is cell cycle dependent or may be a result
of cell proliferation or a secondary event rather than
direct thrombin stimulation. Regardless, these results
suggest that monocytes are the cells involved in the
production of IL6 by PBMC. Unexpectedly, thrombin
did not induce IL8 production by monocytes under the
conditions used for these studies. It is possible that
the IL8 production induced by thrombin at early time
points (Sower et al., 1996) is turned off by the cells by
48 hr or that the IL8 has been degraded. These data
make sense physiologically because in early wounding
the IL8 production induced by thrombin may be necessary to recruit other cells into the wound, but after 48
hr this recruitment may no longer be needed. These
data further implicate thrombin as an important mediator of the inflammatory response.
The mechanism by which thrombin induces enhanced proliferation is not yet understood. High/moderate affinity thrombin receptors, such as those originally
reported on fibroblasts (Rasmussen et al., 1991), have
been identified in a number of cell types (Carney et al.,
1992b) including macrophages (Kudahl et al., 1991),
neutrophils (Bizios et al., 1986), T lymphocytes (Tordai
et al., 1993), and NK cells (Howells et al., 1993). There
is evidence that this high-affinity binding does not correlate with expression of PAR1, suggesting that highaffinity binding may be to a separate N-PAR component. The present study confirms that human monocytes and differentiated U937 cells bind thrombin and
express PAR1. Previous studies have shown that functional thrombin receptors are present on U937 cells
(Joseph and MacDermot, 1993), but there had been no
evidence presented that these cells bind thrombin with
wlcpal
W Liss: JCP
JCP-10284
83
THROMBIN-INDUCED MONOCYTE ACTIVATION
Fig. 6. A: Increase in specific thrombin binding upon U937 differentiation. U937 and IFNg-differentiated U937 cells were incubated with
biotinylated thrombin (hatched bars) or without biotinylated thrombin (unhatched bars) for 90 min as described in Materials and Methods. Thrombin binding was determined by cell sorting using streptavidin-coated magnetic beads. Asterisk indicates statistical significance
(P õ 0.05). Data are representative of two independent experiments
using duplicate samples. B: U937 and IFNg-differentiated U937 cells
were incubated with 125I-thrombin for 90 min { 100-fold excess cold
thrombin and filtered through GF/C filters as described in Materials
and Methods. Specific binding Å total counts - nonspecific counts.
Asterisk indicates statistical significance where P õ 0.05. Data are
representative of three independent experiments using duplicate samples.
high enough affinity to measure in or 125I-thrombin
binding or nonradioactive thrombin binding experiments. Consistent with these findings, in HL60 cells,
monocytic-differentiated lineage appears to correlate
with enhanced high-affinity thrombin binding (BarShavit et al., 1987). Additional experiments are now
underway to establish if the increased thrombin binding observed in our studies could be due to the increased expression of PAR1 or if it is due to the increase
in expression of the N-PAR component.
Physiologically, thrombin enhancement of monocyte
activation could be extremely important in regulating
the wound healing process and subsequent events triggered by release of IL6 and other cytokines (Sporn and
Roberts, 1993; Slavin, 1996). Indeed, IL6, IL1b, and
TNFa are known as fibrogenic cytokines (Kovacs, 1991)
because they promote later wound healing responses.
Thrombin is always present in the wound. Following
clot formation, wound healing proceeds involving a series of events. First, neutrophils, lymphocytes, and
monocytes accumulate into the area of the wound.
Next, fibroblasts and endothelial and epithelial cells
migrate into the wound, where they proliferate. We
have previously shown that thrombin activates T lymphocytes, enhancing T-cell proliferation and IL6 pro-
duction (Naldini et al., 1993). We now show that differentiated monocytes are activated by thrombin to produce IL6, IL1b, and TNFa. In addition, thrombin
induces IL6 production in fibroblasts and IL6 enhances
the thrombin receptor expression in these cells (Sower
et al., 1995). Thus, the ability of thrombin to induce
the production of fibrogenic cytokines (IL6, IL1b, and
TNFa) by monocytes could have important physiological consequences in the orchestration of wound healing
and the inflammatory response than previously recognized.
/ 893f$$0284
07-20-98 20:24:53
ACKNOWLEDGMENTS
The authors thank Dr. F. Carraro and Prof. G.P. Pessina for valuable discussion and Mrs. P. Marrocchesi
for assistance with the manuscript preparation.
LITERATURE CITED
Bar-Shavit, R., Kahn, A., Fenton, J.W.I., and Wilner, G.D. (1983)
Chemotactic response of monocytes to thrombin. J.Cell Biol.,
96:282 – 285.
Bar-Shavit, R., Kahn, A.J., Mann, K.G., and Wilner, G.D. (1986) Identification of a thrombin sequence with growth factor activity on
macrophages. Proc. Natl. Acad. Sci.USA, 83:976 – 980.
Bar-Shavit, R., Hruska, K.A., Kahn, A.J., and Wilner, G.D. (1987)
Thrombin chemotactic stimulation of HL-60 cells: Studies on throm-
wlcpal
W Liss: JCP
JCP-10284
84
NALDINI ET AL.
bin responsiveness as a function of differentiation. J.Cell. Physiol.,
131:255 – 261.
Bizios, R., Lai, L., Fenton, J.W.II., and Malik, A.B. (1986) Thrombininduced chemotaxis and aggregation of neutrophils. J.Cell. Physiol.,
128:485 – 490.
Brass, L.F., Manning, D.R., Williams, A.G., Woolkalis, M.J., and
Poncz, M. (1991) Receptor and G protein-mediated responses to
thrombin in HEL cells. J.Biol.Chem., 266:958 – 965.
Carney, D.H., and Cunningham, D.D. (1978) Cell surface action of
thrombin is sufficient to initiate division of chick cells. Cell, 14:811 –
823.
Carney, D.H., Herbosa, G.J., Stiernberg, J., Bergmann, J.S., Gordon,
E.A., Scott, D., and Fenton, J.W.I. (1986) Double-signal hypothesis
for thrombin initiation of cell proliferation. Semin.Thromb. Hemost., 12:231 – 240.
Carney, D.H., Mann, R., Redin, W.R., Pernia, S.D., Berry, D., Heggers,
J.P., Hayward, P.G., Robson, M.C., Christie, J., Annable, C., Fenton
II, J.W., and Glenn, K.C. (1992a) Enhancement of incisional wound
healing and neovascularization in normal rats by thrombin and
synthetic thrombin receptor-activating peptides. J.Clin. Invest.,
89:1469 – 1477.
Carney, D.H., Redin, W., and McCroskey, L. (1992b) Role of highaffinity thrombin receptors in postclotting cellular effects of thrombin. Semin.Thromb. Hemost., 18:91 – 103.
Clohisy, D.R., Erdmann, J.M., and Wilner, G.D. (1990) Thrombin
binds to murine bone marrow-derived macrophages and enhances
colony-stimulating factor-1-driven mitogenesis. J. Biol. Chem.,
265:7729 – 7732.
Colotta, F., Sciacca, F.L., Sironi, M., Luini, W., Rabiet, M.J., and
Mantovani, A. (1994) Expression of monocyte chemotactic protein1 by monocytes and endothelial cells exposed to thrombin. Am. J.
Pathol., 144:975 – 985.
Gardner, I.D., and Remington, J.S. (1978) Aging and the immune
response. II. Lymphocyte responsiveness and macrophage activation in Toxoplasma gondii-infected mice. J. Immunol., 120:944 – 949.
Gospodarowicz, D., Brown, K.D., Birdwell, C.R., and Zetter, B.R.
(1978) Control of proliferation of human vascular endothelial cells.
Characterization of the response of human umbilical vein endothelial cells to fibroblast growth factor, epidermal growth factor, and
thrombin. J.Cell Biol., 77:774 – 788.
Grand, R.J.A., Turnell, A.S., and Grabham, P.W. (1996) Cellular consequences of thrombin-receptor activation. Biochem.J., 313:353 –
368.
Gurwitz, D., and Cunningham, D.D. (1988) Thrombin modulates and
reverses neuroblastoma neurite outgrowth. Proc. Natl.Acad. Sci.
USA, 85:3440 – 3444.
Harlan, J.M., Thompson, P.J., Ross, R.R., and Bowen-Pope, D.F.
(1986) Alpha thrombin induces release of platelet-derived growth
factor-like molecule(s) by cultured human endothelial cells. J.Cell
Biol., 103:1129 – 1133.
Hoffman, M., and Church, F.C. (1993) Response of blood leukocytes
to thrombin receptor peptides. J. Leukoc. Biol., 54:145 – 151.
Howells, G.L., Macey, M., Curtis, M.A., and Stone, S.R. (1993) Peripheral blood lymphocytes express the platelet-type thrombin receptor.
Br. J. Haematol., 84:156 – 160.
Jenkins, A.L., Howells, G.L., Scott, E., Le Bonniec, B.F., Curtis, M.A.,
and Stone, S.R. (1995) The response to thrombin of human neutrophils: Evidence for two novel receptors. J.Cell Sci., 108:3059 – 3066.
Jones, A., and Geczy, C.L. (1990) Thrombin and factor Xa enhance
the production of interleukin-1. Immunology, 71:236 – 241.
Joseph, S., and MacDermot, J. (1993) The N-terminal thrombin receptor fragment SFLLRN, but not catalytically inactive thrombin-derived agonists, activate U937 human monocytic cells: Evidence for
receptor hydrolysis in thrombin-dependent signalling. Biochem. J.,
290:571 – 577.
Kovacs, E.J. (1991) Fibrogenic cytokines: The role of immune mediators in the development of scar tissue. Immunol.Today, 12:17 – 23.
Kudahl, K., Fisker, S., and Sonne, O. (1991) A thrombin receptor in
resident rat peritoneal macrophages. Exp.Cell Res., 193:45 – 53.
Levin, E.G., Marzec, V., Anderson, J., and Harker, L.A. (1984) Throm-
/ 893f$$0284
07-20-98 20:24:53
bin stimulates tissue plasminogen activator release from cultured
human endothelial cells. J.Clin. Invest., 74:1988 – 1995.
Mosmann, T. (1983) Rapid colorimetric assay for cellular growth and
survival: Application to proliferation and cytotoxicity assays. J. Immunol. Method., 65:55 – 63.
Naldini, A., and Carney, D.H. (1996) Thrombin modulation of natural
killer activity in human peripheral lymphocytes. Cell. Immunol.,
172:35 – 42.
Naldini, A., Carney, D.H., Bocci, V., Klimpel, K.D., Asuncion, M.,
Soares, L.E., and Klimpel, G.R. (1993) Thrombin enhances T cell
proliferative responses and cytokine production. Cell. Immunol.,
147:367 – 377.
Naldini, A., Carraro, F., Silvestri, S., and Bocci, V. (1997) Hypoxia
affects cytokine production and proliferative responses by human
peripheral mononuclear cells. J.Cell. Physiol., 173:335 – 342.
Oberg, F., Larsson, L.-G., Anton, R., and Nilsson, K. (1991) Interferon
gamma abrogates the differentiation block in v-myc-expressing U937 monoblasts. Proc. Natl. Acad. Sci.USA, 88:5567 – 5571.
Rasmussen, U.B., Craviari, V.V., Jallat, S., Schlesinger, Y., Pages, G.,
Pavirani, A., Lecocq, J.-P., Pouyssegur, J., and Van ObberghenShilling, E. (1991) DNA cloning and expression of a hamster athrombin receptor coupled to Ca2/ mobilization. FEBS Lett.,
288:123 – 128.
Roberts, P.J., Devalia, V., Faint, R., Pizzey, A., Bainton, A.L., Thomas,
N.S.B., Pilkington, G.R., and Linch, D.C. (1991) Differentiationlinked activation of the respiratory burst in a monocytic cell line
(U937) via Fc tRII. A study of activation pathways and their regulation. J.Immunol., 147:3104 – 3115.
Shuman, M.A. (1986) Thrombin-cellular interactions. Ann. N.Y. Acad.
Sci., 485:228 – 238.
Slavin, J. (1996) The role of cytokines in wound healing. J. Pathol.,
178:5 – 10.
Sower, L.E., Froelich, C.J., Carney, D.H., Fenton, J.W., and Klimpel,
G.R. (1995) Thrombin induces IL-6 production in fibroblasts and
epithelial cells - Evidence for the involvement of the seven-transmembrane domain (STD) receptor for alpha-thrombin. J. Immunol,
155:895 – 901.
Sower, L.E., Froelich, C.J., Allegretto, N., Rose, P.M., Hanna, W.D.,
and Klimpel, G.R. (1996) Extracellular activities of human granzyme A - Monocyte activation by granzyme A versus alpha-thrombin. J. Immunol., 156:2585 – 2590.
Sporn, M.B., and Roberts, A.B. (1993) A major advance in the use of
growth factors to enhance wound healing. J.Clin. Invest., 92 :2565 –
2566.
Stern, D.M., Bank, I., Nawroth, P.P., Cassimeris, J., Kisiel, W., Fenton, J.W.I., Dinarello, C., Chess, L., and Jaffe, E.A. (1985) Selfregulation of procoagulant events on the endothelial cell surface. J.
Exp. Med., 162:1223 – 1235.
Sundström, C., and Nilsson, K. (1976) Establishment and characterization of a human histiocytic lymphoma cell line. Int. J. Cancer,
17:565 – 577.
Tordai, A., Fenton, J.W.I., Andersen, T., and Gelfand, E.W. (1993)
Functional thrombin receptors on human T lymphoblastoid cells.
J. Immunol., 150:4876 – 4886.
Ucla, C., Roux-Lombard, P., Fey, S., Dayer, J.-M., and Mach, B. (1990)
Interferon gamma drastically modifies the regulation of interleukin
1 genes by endotoxin in U937 cells. J. Clin. Invest., 85:185 – 191.
Van Snick, J. (1990) Interleukin-6: An overview. Annu. Rev. Immunol., 8:253 – 278.
Vouret-Craviari, V., Van Obberghen Schilling, E., Rasmussen, U.B.,
Pavirani, A., Lecocq, J.-P., and Pouyssegur, J. (1992) Synthetic
alpha-thrombin receptor peptides activate G protein-coupled signaling pathways but are unable to induce mitogenesis. Mol. Biol.Cell,
3:95 – 102.
Weiss, R.H., and Nuccitelli, R. (1992) Inhibition of tyrosine phosphorylation prevents thrombin-induced mitogenesis, but not intracellular
free calcium release, in vascular smooth muscle cells. J. Biol. Chem.,
267:5608 – 5613.
Weksler, B., Ley, C.W., and Jaffe, E.A. (1978) Stimulation of endothelial cell prostacyclin production by thrombin, trypsin, and the ionophore A23187. J.Clin. Invest., 62:923 – 930.
wlcpal
W Liss: JCP
JCP-10284