close

Вход

Забыли?

вход по аккаунту

?

Superoxide anion production and phagocytosis of crystals by cultured endothelial cells.

код для вставкиСкачать
105
SUPEROXIDE ANION PRODUCTION AND
PHAGOCYTOSIS OF CRYSTALS BY CULTURED
ENDOTHELIAL CELLS
GERALD F. FALASCA, ANAND RAMACHANDRULA, KATHLEEN A. KELLEY,
CAROLYN R. O’CONNOR, and ANTONIO J. REGINATO
Objective. To show that cultured human umbilical vein endothelial cells (HUVEC) are capable of
phagocytizing inflammation-causing crystals and of generating superoxide anion (SOA) during phagocytosis.
Methods. The superoxide dismutase-inhibitable
reduction of nitroblue tetrazolium (NBT) dye was used
as a measure of SOA production. Phagocytosis was
quantified by light microscopy and confirmed by transmission electron microscopy. Cytochrome C was also
studied but was found to undergo spontaneous reduction
by monosodium urate (MSU) without cells.
Results. Crystals of MSU, calcium oxalate, hydroxyapatite, and calcium pyrophosphate dihydrate
(CPPD) were phagocytized and, except for the CPPD
crystals, induced NBT reduction. Cholesterol and choPresented in part at the 53rd Annual Meeting of the American College of Rheumatology, Cincinnati, OH, June 1989.
From the Division of Rheumatology, Department of Medicine, Cooper HospitalKJniversity Medical Center, Robert Wood
Johnson Medical School at Camden, University of Medicine and
Dentistry of New Jersey, Camden, New Jersey.
Dr. O’Connor’s work was supported by grant 12-89 from
the Foundation of the University of Medicine and Dentistry of New
Jersey (UMDNJ). Dr. Falasca’s work was supported by a faculty
practice research support grant from UMDNJ. Research funds were
provided by the Department of Medicine of Cooper Hospital/
University Medical Center under the direction of Edward D. Viner,
MD.
Gerald F. Falasca, MD: Assistant Professor of Medicine;
Anand Ramachandrula, PhD: Director, Department of Medicine
Laboratory; Kathleen A. Kelley, BS: Research Assistant; Carolyn
R. O’Connor, MD: Associate Professor of Clinical Medicine; Antonio J. Reginato, MD: Professor of Medicine.
Address reprint requests to Gerald F. Falasca, MD, Department of Medicine, Education and Research Building, Cooper Hospitalluniversity Medical Center, 401 Haddon Avenue, Camden, NJ
08103.
Submitted for publication February 22, 1991; accepted in
revised form September 17, 1992.
Arthritis and Rheumatism, Vol. 36, No. 1 (January 1993)
lesterol monohydrate were neither phagocytized nor did
they induce NBT reduction.
Conclusions. Endothelial cells may be a significant source of oxygen radicals in crystal-associated and
other arthritides.
The spectrum of crystal-associated inflammation is broad, and is perhaps most familiar clinically to
rheumatologists in the form of acute crystalline arthropathy due to monosodium urate (MSU),calcium pyrophosphate dihydrate (CPPD), or hydroxyapatite
(HA). Less common clinical entities include the
crystal-associated vasculopathies, as observed in primary and secondary oxalosis (1-5), ethylene glycol
poisoning (6), and calciphylaxis due to chronic renal
failure and dialysis (7-1 l ) , as well as cholesterol
embolization syndrome (12), cryoglobulin crystal vasculopathy (13), vascular calcification seen in diabetics
(14), and familial forms of hydroxyapatite deposition
diseases (15). Another clinical example of interaction
between crystals and endothelium is in the formation
and ulceration of atherosclerotic plaques which contain cholesterol and HA crystals and sometimes MSU
crystals (16,17).
Common to all these disease entities is an
interaction of crystals with various tissue cells. The
relationship between crystals and endothelial cells has
only rarely been investigated, despite the fact that
vascular endothelium almost certainly plays an important role in the development of acute arthritis. Synovial effusions are filtrates of plasma which pass
through fenestrations in subsynovial capillary endothelial cells in response to dynamic changes in endothelial permeability (18). It has been shown that crystals injected into the joints of humans and of dogs
FALASCA ET AL
106
Table 1.
Physical characteristics of crystals studied
Average
length
Range of
length
Crystal
(w)
(w)
Monosodium urate
Hydroxyapatite
Calcium pyrophosphate dihydrate
Calcium oxalate
Cholesterol
Cholesterol monohydrate
4
3
12
5
6
6
cause the formation of large synovial effusions in 2 or
3 hours, presumably due to direct or indirect effects on
the permeability of the subsynovial endothelial cells to
plasma constituents (19). Furthermore, endothelial
cells have been shown to be capable of generating
superoxide anion (SOA) (20,21) and of being phagocytic (22-24).
Because of these features, we studied the interaction between various crystals and cultured endothelial cells. Our goal was to investigate whether endothelial cells phagocytize crystals of monosodium
urate, calcium pyrophosphate dihydrate, hydroxyapatite, calcium oxalate (COX), cholesterol, and cholesterol monohydrate (CM), and to determine whether
this process generates superoxide anion.
MATERIALS AND METHODS
Crystals. MSU was prepared using the method of
Seegmiller et al (25). CPPD was purchased from Dr. K. P.
Pritzger (Beth Israel Hospital, Toronto, Ontario, Canada).
HA (type 1) was purchased from Sigma (St. Louis, MO), and
calcium oxalate was purchased from Fisher Scientific
(Springfield, NJ). Cholesterol monohydrate was prepared
from cholesterol (Sigma) according to the method of Hasselbacher and Hahn (26). Crystals were ground to appropriate
sizes with a ceramic mortar and pestle, then heated (except
cholesterol and CM) to 180°C for 3 hours to destroy endotoxin. Aliquots of crystals were suspended in freshly prepared Krebs-Ringer phosphate buffer with dextrose (KRPD)
by vortex mixing. Cholesterol and CM crystals were further
suspended by sonication in polystyrene test tubes using a
sonicator (model W-220F; Heat Systems-Ultrasonics, Inc.,
Farmingdale, NY) with a microtip at 70% power for 45
seconds. The concentration of the crystals in suspension was
calculated by counting crystals in a hemocytometer (Table
1 ).
Endothelial cells. Endothelial cells were isolated from
human umbilical veins as described elsewhere (27), but with
minor modifications. Briefly, umbilical veins were cannulated, washed with Hanks' balanced salt solution (HBSS)
(calcium- and magnesium-free), and filled with HBSS containing type I collagenase, 1.0 mg/ml (Sigma). After a 25-
1-16
1-28
1-3 1
1-25
1-30
1-20
No. of
crystals per
0.4 mg/ml
( x I09
Crystal shape
1.80
0.53
1.70
0.24
0.07
0.07
Small needles
Chunks
Long needles
Chunks
Plates and chunks
Plates and chunks
minute incubation at room temperature, the detached cells
were obtained by washing with MI99 (Gibco, Grand Island,
NY) containing 10% iron-supplemented calf serum (Hyclone, Logan, UT), then resuspended in complete medium
consisting of M199, 10% heat-inactivated fetal calf serum
(Hyclone), 15% heat-inactivated pooled human serum, endothelial cell growth factor (28), 2 mM L-glutamine, 0.02M
HEPES, 200 units/ml penicillin, and 200 pg/ml streptomycin
(Sigma).
The cells were cultured in 25-cmZ tissue culture
flasks (Falcon Labware; Becton Dickinson, Lincoln Park,
NJ) coated with 1% autoclaved gelatin at 37"C, in humidified
air with 5% CO,. Third-generation cells from single cords
were used throughout each experiment. For the final passage, cells were seeded into 24-well plates (Falcon) coated
with 8 pg of fibronectin (Sigma) per well at a concentration
of I to 2 x lo5 cells/ml, with 1 .O ml of complete medium per
well. Cells were cultured to confluence (3 to 5 x lo5
cells/well) for all experiments.
Cells were confirmed as endothelial cells by their
characteristic morphology and ability to bind Factor VIII
antibodies in an indirect immunofluorescence assay. Contamination with smooth muscle cells was less than 1% in the
third generation, as assessed by immunofluorescence using
anti-actin antibodies (Sigma). Crystals, media, and reagents
were tested for endotoxin contamination with the limulus
amebocyte assay and were found to contain less than 10
pg/ml. Viability of cells, as assessed by trypan blue exclusion, was not affected by addition of crystals, nitroblue
tetrazolium (NBT), superoxide dismutase (SOD), or cytochalasin B. Viability was also confirmed using a lactic
dehydrogenase (LDH) release assay (29). Endothelial cells
(duplicate samples in 24-well plates) were incubated with
HBSS, with or without crystals (0.4 mg/ml), for 2 hours.
Viability was calculated as:
% viable cells =
(
I -
LDH release using test substance
LDH release using Triton X-100
)
x 1
0
m
Results were 90.9% for HBSS alone, 99.3% for MSU, 95.9%
for HA, 100% for CPPD, 89.3% for COX, and 95.4% for CM.
Light microscopy. Crystals were suspended in KRPD
at 25°C and added to wells containing endothelial cells that
had been washed with KRPD. These were incubated in
humidified air with 5% CO, at 37°C. At specified times, the
107
SUPEROXIDE PRODUCTION BY ENDOTHELIAL CELLS
Table 2.
Superoxide anion production by endothelial cells stimulated with concanavalin A (Con A)*
Optical density at 700 nm
Stimulant
None (HBSS)
Con A
5 @ml
25 &ml
50 @ml
MSU, 0.4 mg/ml
HA, 0.4 mg/ml
Without SOD
With SOD
0.0598 C 0.0036
0.0161
0.0696 f 0.0037
0.0797 f 0.0048
0.0697 ? 0.0034
0.1015 f 0.0231
0.1122 2 0.0399
0.0230 ? 0.0017
0.0282 f 0.0020
0.0419 f 0.0042
0.0556 2 0.0292
0.0862 2 0.0476
P
-
f 0.0048
0.0290
0.0046
0.0258
0.0118
0.0398
* Confluent endothelial cell monlayers were incubated at 37°C for 90 minutes with Hanks’balanced
salt solution (HBSS) with Con A or crystal suspensions, with or without superoxide dismutase (SOD)
100 &ml. Plates were processed as described in Materials and Methods. Values are the mean f SD
of 3 experiments performed in duplicate. P values are for stimulated condition without SOD versus
HBSS (no stimulant) without SOD. MSU = monosodium urate; HA = hydroxyapatite.
crystal suspensions were aspirated, and 0.5 ml of 0.15M
sodium fluoride was added for 1 minute to stop phagocytosis
(30). Cells were washed with 1.0 ml of KRPD to remove
excess crystals. Cells were detached with 1.0 ml of trypsin
solution (0.05% trypsin and 0.02% EDTA; Sigma). A hemocytometer was used to count intracellular crystals under
regular light at I ,OOOx magnification; 70-80 cells in each of 2
aliquots from duplicate wells at each time period were
counted. Crystals which appeared to be adherent to the
outside of the cell membrane were not counted.
The degree of granularity of normal cells without
crystals was assessed at the beginning of each counting session
to control for intracellular granules that might resemble crystals. Cell lysis was determined by counting 4 aliquots of cells
in a hemocytometer before and after each experiment; lysis
was always less than 10%. Viability was greater than 85%, as
determined by exclusion of 0.5% trypan blue dye.
For some experiments, cells were preincubated with
5 pg/ml of cytochalasin B to prevent phagocytosis. Phagocytosis was inhibited in preliminary experiments by incubating cells and crystals at 4°C. The results using this method
did not differ significantly from those using cytochalasin B,
as assessed by light microscopy. The cytochalasin B method
was therefore chosen for convenience.
Nitroblue tetrazolium assay. NBT dye (1.0 mg/ml,
grade 3; Sigma) was dissolved in KRPD and filtered. Crystals, with or without SOD or cytochalasin B (both from
Sigma), were suspended in the NBT solution, and 1.0-ml
aliquots were added to wells containing cells. Concanavalin
A (Con A) (5-50 pg/ml final concentration; Sigma) was
added to some wells. For experiments utilizing cytochalasin
B, cells were pretreated for 45 minutes with growth medium
containing cytochalasin B at 5 pg/ml, the same concentration
used in the experiments. Plates were incubated at 37°C in
humidified air with 5% COz. At specified times, the crystal
suspensions were aspirated and discarded.
The reaction was stopped by the addition of 1.0 ml of
70% methanol, and the plates were processed according to
the method of Rook et al (31), with minor modifications.
Briefly, plates were washed with 4 changes of 100% methanol to remove unreduced NBT. The reduced formazan
precipitate remained visible as adherent purple granules on
the bottom of the wells. After air drying the plates, 0.75 ml
of 2M potassium hydroxide was added to dissolve cell
membranes. The formazan was then solubilized by the
addition of 0.875 ml of dimethyl sulfoxide to the KOH.
Specimens were centrifuged 4 minutes at 15,000 revolutions
per minute in a Beckman E Microfuge and read at 700 nm
(OD [optical density] 700) on a Beckman DU-7 spectrophotometer against a blank cuvette containing KOH and DMSO.
Optical densities obtained using cell-free control wells containing stimuli were then subtracted from the OD values
obtained from wells containing cells. Cell-free control wells
incubated for 90 minutes with 0.4 mg/ml of crystal suspensions yielded mean 2 SD OD values of 0.0032 2 0.0020 for
buffer alone, 0.0047 5 0.0007 for MSU, 0.0027 zt 0.0002 for
HA, 0.0035 2 0.0010 for CPPD, 0.0073 0.0031 for COX,
0.0088 2 0.0016for cholesterol, and 0.0083 zt 0.0002 for CM.
For comparison, wells containing cells and buffer without
crystals (“baseline conditions”) typically yielded OD values
in the range of 0.0400 to 0.1200.
Cytochrome C assay. Cytochrome C (type 3 or type 6;
Sigma) 0.99 mg/ml was dissolved in KRPD. Crystals, SOD,
or KRPD were added just prior to use. Aliquots (1 .O ml) of
crystal suspension were added to wells and incubated at
37°C for 60 minutes, as described above. Supernatants were
aspirated and centrifuged (as above) at 4°C. Optical density
at 550 nm was read immediately against a blank consisting of
incubated cytochrome C solution alone.
Statistical analysis. Data are reported as the mean
SD. Except as otherwise stated, Student’s 2-tailed t-test for
paired or for pooled data was used to compare differences of
means, with significance at P < 0.05.
*
*
RESULTS
Superoxide production by endothelial cells. Reduction of NBT was used as a n indicator of SOA
production (32-35). Con A, a known stimulant of
superoxide anion production in neutrophils (34), was
FALASCA ET AL
108
300
0
H
250
I-
2 200
150
100
~
50
n
CNT 0,050.1 0 . 2 0 . 4 0 . 8 1 . 6 6 . 4
CNTO.050.1 0 . 2 0 . 4 0 . 8 1 . 6 3 . 2 6 . 4
MONOSODIUM URATE
t
2 3501
v
W T O . 0 5 0 . 1 0 . 2 0 . 4 0 . 8 1.6 3 . 2 6.4
HYDROXYAPATITE
CALCIUM PYROPHOSPHATE
t
CNT0.05010.20.40.8 1.63.26.4
CN10.050.1 0 . 2 0.4 1 . 6 3 . 2 6.4
CNT0.050.10.20.40.8l.63.26.4
CALCIUM OXALATE
CHOLESTEROL
CHOLEST. MONOHYDRATE
Figure 1. Dose-response histograms for nitroblue tetrazolium (NBT) reduction by endothelial cells as
a function of crystal concentration. Cell monolayers were incubated for 90 minutes with crystal
suspensions (0.05-6.4 mg/ml) in Krebs-Ringerphosphate buffer containing 1 .O mg/rnl NBT dye. Values
are the mean and SD. CNT = control; cholest. = cholesterol.
used to compare the magnitude of SOA production in
response to crystals (Table 2). HA and MSU (0.4
mg/ml) were approximately as active as Con A (25
pg/ml) in stimulating SOA production.
Dose-response curves plotting NBT reduction
as a function of crystal concentration are shown in
Figure 1. MSU, H A , and COX all exhibited marked
dose-response phenomena, while CPPD, cholesterol,
and CM stimulated little NBT reduction above baseline, even at high doses. There was no significant SOA
production above baseline by cells incubated with
crystal suspensions which had been rendered crystalfree by centrifugation (data not shown).
Next, we sought to compare crystal activity on
a per-crystal basis. To do this, crystal suspensions
containing approximately equal numbers of crystals
per milliliter were used. The results and the concentration (in mg/ml) of each crystal type required to give
approximately 1.8 x lo8 crystals/ml are shown in
Figure 2. H A , COX, and MSU crystals were more
active, on a per-crystal basis, in reducing NBT than
were CPPD, cholesterol, or CM crystals.
To rule out an active inhibitory effect of CPPD
on NBT reduction as the cause of the relative inactivity of CPPD compared with MSU, both types of
crystals were suspended together at a concentration of
0.4 mg/ml and compared with MSU alone at 0.4 mg/ml.
There was no difference in the amount of NBT reduced by the mixture compared with that reduced by
MSU alone (data not shown), suggesting that CPPD
does not actively inhibit SOA production.
For the following experiments MSU, H A , and
CPPD were studied in detail. The time courses of NBT
reduction by endothelial cells stimulated with MSU,
HA, or CPPD are shown in Figure 3. By 60 minutes,
cells stimulated with MSU and H A reduced significantly more NBT than did unstimulated cells, and by
90 minutes, reduction was significant for MSU, CPPD,
and H A .
The progressive increase in NBT reduction
over time by baseline cells incubated with buffer alone
may reflect a response to the perturbation of the
culture environment caused by the absence of serum
in the buffer, since the results of preliminary experi-
109
SUPEROXIDE PRODUCTION BY ENDOTHELIAL CELLS
-
350
+ I
I
Z
*
Monosodium urate
300
n
.351 1
I
x
*
W
T
0
0 Hydroxyapatite
n
A Calcium pyrophosphate
0 Control (buffer)
0
v
Z
z
0
0
H
.20
.15
H
*
T
W
L
r
m
T
t-
=
100
01’
m
z
50
BUFFER MSU
HA CPPD COX CHOL
CM
0 . 4 1 1.35 0 . 4 2 3.15 9.75 9 . 7 5
I
0
30
60
90
TIME (minutes)
120
Figure 3. Kinetics of nitroblue tetrazolium (NBT) reduction by
endothelial cells stimulated with crystals. Cell monolayers were
incubated with crystal suspensions (0.4 mglml in Krebs-Ringer
phosphate buffer) containing 1.0 mglml NBT dye. Values are the
mean ? SD of 3 experiments performed in duplicate. * = P < 0.05
versus buffer at the time point indicated. O.D. = optical density.
CRYSTAL TYPE AND CONCENTRATION
(mg/ml)
Figure 2. Nitroblue tetrazolium (NBT) reduction by equal numbers
of crystals per milliliter. Cell monolayers were incubated with
crystal suspensions in Krebs-Ringer phosphate buffer with 1.O
mglml NBT dye for 90 minutes at 37°C. The crystal concentrations
(mglml) were those utilized to give 1.8 x 10* crystalslml for each
crystal type. Values are the mean and SD of 3 experiments performed in duplicate. * = P < 0.05 versus buffer alone. MSU =
monosodium urate; HA = hydroxyapatite; CPPD = calcium pyrophosphate dihydrate; COX = calcium oxalate; CHOL = cholesterol; CM = cholesterol monohydrate.
ments performed in our laboratory show that inclusion
of serum in the buffer almost eliminates baseline NBT
reduction, with minimal effects on crystal-induced
NBT reduction (data not shown). The phenomenon of
“baseline” SOA production has been reported by
others using serum-free reaction buffer (20,32).
Viable cells were required for NBT reduction in
our system. There was very little NBT reduction in
control wells containing crystals and NBT but no cells
(OD 700 nm <0.01), or in wells in which cells had been
incubated at 56°C for 40 minutes prior to use (OD 700
nm c0.02).
Effect of superoxide dismutase. The effect of
SOD on NBT reduction is shown in Figure 4 for MSU,
CPPD, and HA stimuli. Modest decrements in NBT
reduction were seen in each case, indicating that SOA
was present. The degree of inhibition is similar to that
obtained by other investigators using NBT (31), but it
does not approach 100% inhibition, since the large
SOD molecule is excluded from the intracellular
space, where most of the NBT is believed to be
reduced (32) and there may be some nonspecific NBT
reduction (31). There was no significant inhibition of
SOA production using SOD which had been heatinactivated for 10 minutes at 100°C (data not shown).
Phagocytosis of crystals. All of the crystal types
tested, except cholesterol and CM, were readily
200
&?
v
z
150
I0
3
f2 100
rY
*
m
Z
50
0
BUFFER
n = 7
MSU
HA
n = 7
n = 4
CPPD
n=3
Figure 4. Effect of superoxide dismutase (SOD) on nitroblue tetrazolium (NBT) reduction. Endothelial cells were incubated at 37°C
for 90 minutes with crystal suspensions (0.4 mglml) in Krebs-Ringer
phosphate buffer with 1.0 mdml NBT dye, with or without SOD.
Values are the mean and SD of n experiments (shown across the
bottom). * = P < 0.05. See Figure 2 for other definitions.
FALASCA ET AL
110
phagocytized by endothelial cells, as assessed by both
light and electron microscopy. The percentage of cells
phagocytizing at least one crystal and the average
number of crystals phagocytized (for cells phagocytizing at least one crystal) at sequential intervals are
shown in Figure 5 for MSU, CPPD, and HA. At 90
minutes of incubation, the mean ? SD number of
intracellular crystals was quantified for all 6 crystal
types: 14.1 k 1.5 for MSU, 4.6 & 0.4 for CPPD, 2.4 +0.8 for HA, 6.7 -+ 0.6 for COX, 1.3 ? 0.2 for cholesterol, and 1.4 k 0.2 for CM. The mean number of
crystals per cell was consistently greater for MSU than
for the other crystals. MSU crystals appeared to
adhere rapidly to the surface of the endothelial cells,
which may account for the more extensive degree of
phagocytosis compared with that of the other crystals.
Cholesterol and CM crystals were the least
efficiently phagocytized. Only a few of these crystals
appeared to be intracellular when examined by light
microscopy, and in contrast to other crystal types, no
cholesterol or CM crystals were found intracellularly
when evaluated by transmission electron microscopy.
Light photomicrographs of endothelial cells in the
process of phagocytizing crystals are shown in Figure
6. The extent of phagocytosis leveled off after 1.5-2.0
hours; however, if incubation continued at 37°C for 7
or 8 hours, the crystals moved into a circular or
spherical distribution within the cells (Figure 6B).
Effect of cytochalasin B. Addition of cytochalasin B resulted in marked decreases in the extent of
phagocytosis of the 3 crystal types studied, as shown
in Table 3. Cytochalasin B did not completely inhibit
phagocytosis, especially in the case of MSU. Table 3
also shows that cytochalasin B significantly inhibited
NBT reduction by cells stimulated with MSU and HA,
but not with CPPD. Cytochalasin B slightly inhibited
NBT reduction with buffer alone; this may have been
due to inhibition of phagocytosis of insoluble, reduced
NBT granules formed during incubation.
To rule out a nonspecific toxic inhibitory effect
of cytochalasin B, some wells were treated with menadione, a vitamin K analog which stimulates radical
production by a nonphagocytic mechanism (21). Cytochalasin B did not inhibit NBT reduction by menadione-stimulated cells (data not shown). Menadione
did not reduce NBT in the absence of cells, or when
using cells that had been incubated at 56°C for 40
minutes prior to use.
Cytochrome C reduction. SOA production by
phagocytic cells has been quantified by other investigators using the method of cytochrome C reduction
n
-0
W
v,
-1
U
I-
m
>
Ly
0
c3
iN
5
I 4
-I
>
0
0
c3
a
I
a
m
J
J
W
0
"
0
30
90
60
120
-I
-I
w
-
0
[Y
MSU
W
a
-
v,
-I
a
I-
v,
>-
-
[Y
0
LL
0
CPPD
-
[Y
w
m
z
3
Z
HA
0'
0
I
I
30
I
I
60
90
TIME ( m i n , )
I
120
Figure 5. Percentage of cells phagocytizing at least one crystal (A)
and average number of crystals phagocytized per cell (B), at
sequential time intervals. Cell monolayers were incubated with
crystal suspensions (0.4 mg/ml) for the periods indicated, then were
detached with trypsin, and studied by light microscopy (see Materials and Methods). Values are the mean SD of 4 samples at each
time point. See Figure 2 for definitions of abbreviations.
*
(36-38); however, we found that MSU (heated or
unheated) resulted in the spontaneous reduction of
cytochrome C in the presence of all buffers tested
(KRPD, HBSS, phosphate buffered saline, and physiologic saline solution). The magnitude of this reduc-
Figure 6. Light micrographs of endothelial cells phagocytizing crystals. Endothelial cell monolayers were incubated with crystal suspensions
in Krebs-Ringer phosphate buffer, detached with trypsin, and photographed. A, Monosodium urate (MSU) crystals at 60 minutes. B, MSU
crystals at 8 hours. The crystals have formed a ring-like structure within the cell (arrowheads). C, Calcium pyrophosphate dihydrate crystals
at 60 minutes. D, Hydroxyapatite at 60 minutes. (Original magnification x 1,000 for A and C, and X 400 for B and D.)
Table 3. Effect of cytochalasin B on nitroblue tetrazolium (NBT) reduction and phagocytosis*
NBT reduction
(% of control)
Crystal
Buffer
Monosodium urate
Hydroxyapatite
Calcium pyrophosphate
dihydrate
Phagoc ytosis
(no.of crystalskell)
Without
CB
With CB
P
100
-C 44.5
-C 107
109 2 2.3
72.4 2 25.1
142 2 17.7
148 f 36.0
I01
5.5
X0.05
0.05
214
238
<0.05
"
NS
Without
CB
13.7
0.9
2
1.42
2
0.20
4.6 2 0.42
With CB
P
-
-
5.35 +- 1.80
0.20 2 0.07
0.26 t 0.04
<0.01
<0.01
<0.001
* Incubations were performed for 90 minutes, at 37"C, on cells preincubated with cytochalasin B (CB) (see Materials and Methods for details).
NBT reduction is the mean 2 SD of 4 experiments performed in duplicate. Phagocytosis is the mean 2 SD number phagocytized per cell
(averaged over all cells counted, including those without crystals) in 4 experiments performed in duplicate. P values determined by I-tailed
t-test for paired data. - = not applicable; NS = not significant.
111
FALASCA ET AL
112
Table 4. Cytochrome C reduction by urate ion*
Experimental condition
EC absent
No crystals
Xanthine crystals
UA crystals
MSU crystals
MSU crystals + SOD
EC present
MSU crystals
MSU crystals + SOD
OD at 550 nm
P
0.010 2 0.015
0.005 0.009
0.293 f 0.075
0.225 0.068
NSt
<O.O5t
<0.05t
0.161 f 0.061
<0.05$
0.189 f 0.072
0.149 f 0.037
<0.05$
*
*
-
NSS
* Crystals (concentration 0.4 mg/ml) and cytochrome C solution
were incubated 60 minutes at 37°C in 24-well plastic plates, with or
without endothelial cells (EC). Where used, I 0 0 I/.g/ml of superoxide
dismutase (SOD) was added to cytochrome C solution just prior to
mixing with crystals. The optical density (OD) values are the mean
5 SD of 3 experiments performed in duplicate. - = not applicable;
NS = not significant; UA = uric acid.
t Versus control (no EC, crystals, or SOD).
S Versus monosodium urate (MSU) crystals without EC and without SOD.
tion was similar in the presence and in the absence of
cells, and was only partially inhibited by SOD (Table
4). In the absence of cells, uric acid in solution
produced reduction of cytochrome C similar to that
seen with MSU crystals. Neither xanthine, HA,
CPPD, COX, cholesterol, nor CM reduced cytochrome C in the absence of cells.
Transmission electron microscopy studies. The
interaction of each of the 6 crystals with endothelial
cells was studied by transmission electron microscopy
(Figure 7). Phagocytosis was a rapid event: Crystals
were found intracellularly after as little as 15 minutes
(Figure 7C). CPPD and HA crystals almost always
appeared inside phagocytic vacuoles (Figures 7B and
C), whereas MSU crystals always appeared free in the
cytoplasm, a finding previously reported in neutrophils
(29,39). COX crystals were variable in this respect,
only sometimes appearing within vacuoles. An early
phase of phagocytosis seemed to involve finger-like
projections of cytoplasm which surrounded crystals
(Figure 7C). Cholesterol and CM crystals could not be
found intracellularly by transmission electron microscopy. This may represent a sampling error, or it may
be that rare crystals appearing to have been phagocytized under light microscopy were merely adhering to
the exterior cell membrane.
DISCUSSION
We have shown that cultured human umbilical
vein endothelial cells interact with crystals, resulting
in phagocytosis and superoxide anion production. The
latter occurs in a dose-dependent manner for each
crystal type. Some crystals (MSU, HA, and COX)
seem to be more active per crystal, giving support to
the concept that different crystals produce inflammation by different mechanisms. For instance, CPPD was
readily phagocytized, yet resulted in relatively little
NBT reduction compared with crystals such as HA,
COX, and MSU. Previous reports have also suggested
differences in the mechanism of inflammation caused
by CPPD. Di Giovine et al (40) found that MSU, but
not CPPD, stimulated interleukin- 1 release from
mononuclear cells, while Poubelle et a1 (41) found
marked differences between MSU and CPPD with
respect to their ability to activate the excitationresponse coupling sequence of neutrophils. Schumacher et al(39) found that CPPD, but not MSU, crystals
were enclosed in cytoplasmic vacuoles after being
phagocytized by neutrophils. Similarly, we found that
CPPD and HA, but not MSU, were enclosed in
cytoplasmic vacuoles in endothelial cells. This phenomenon alone does not seem to account for differences in NBT reduction among crystal types since HA
was also universally found inside vacuoles, was less
phagocytized than CPPD, and yet reduced more NBT
than did CPPD.
Cytochalasin B had less pronounced effects on
phagocytosis and NBT reduction by MSU crystals
than by the other crystals tested. This may have been
due to the unique propensity of MSU to adhere to the
endothelial cell surface. Adherence alone may be
sufficient to stimulate the respiratory burst. Abramson
et al (36) and Simchowitz et al (32) found significant
SOA release and LDH release, respectively, from
cytochalasin B-treated neutrophils stimulated with
MSU. The importance of cell surface-MSU crystal
interactions has been reviewed elsewhere (42). It has
previously been shown that the respiratory burst can
be stimulated by nonphagocytic mechanisms (35,
42,43). Since cytochalasin B did not completely abolish phagocytosis in the endothelial cells in our experiments, it is difficult to determine if mere adherence of
crystals to the endothelial cell surface is a sufficient
stimulus for SOA production. It appears, based on our
results with CPPD, that phagocytosis is not invariably
associated with SOA production.
The phenomenon of phagocytosis by endothelial cells has been noted previously. Schumacher
found that charcoal particles injected intravenously
appeared inside phagosomes of the subsynovial endothelial cells of monkeys in as little as 30 minutes after
SUPEROXIDE PRODUCTION BY ENDOTHELIAL CELLS
113
Figure 7. Transmission electron micrographs of endothelial cells containing phagocytized crystals. A, Monosodium
urate clefts (arrows), at 60 minutes. B, Hydroxyapatite crystals at 60 minutes. Note vacuolar membrane (arrows)
enclosing crystal clumps. C, Calcium pyrophosphate dihydrate crystals at 15 minutes. Note the presence of vacuolar
membranes (arrows) and finger-like projections (arrowhead)engulfing crystals. (Magnification x 11 ,OOO for A and B;
original magnification x 16,000 for C.)
injection (44,45). Similar findings in the muscle microvasculature of rats have been reported (46). More
recently, there have been reports of phagocytosis of
StuphyZococcus aureus by bovine aorta endothelial
cells (24) and of latex particles by pulmonary endothelium (22), liver endothelial cells (23), and others (47).
Crystal-endothelial cell interaction may play a
role in the pathogenesis of acute crystal-induced in-
114
flammation. Superoxide anion produced as a result of
phagocytosis of crystals by endothelial cells may lead
to increased endothelial permeability and the resulting
rapid accumulation of joint fluid (48).
In addition to their proinflammatory properties,
endothelial cells may also be important in terminating
the acute attack of gout, since SOA has been reported
to dissolve MSU crystals in vitro (49). Superoxide
dismutase has also been suggested as a mediator of
acute gout termination (49), and endothelial cells have
been reported to possess SOD activity (50). Endothelial cells may thus be important in the disposition of
subsynovial deposits of MSU crystals.
The relative contribution to SOA production by
endothelial cells as compared with traditional phagocytes such as neutrophils in crystal-associated inflammation is unknown. Preliminary comparative experiments in our laboratory show that on a per-cell basis,
endothelial cells reduce more NBT than do synovial
cells and neutrophils (data not shown). Reports by
others have shown that synovial cells ( 3 3 3 l), chondrocytes (52), and endothelial cells (20,21) are capable
of producing oxygen radicals and/or hydrogen peroxide. The relative contribution of these cell types to the
total production of oxygen radicals is unknown. Evidence that their contribution may be substantial is
provided indirectly by clinical trials showing symptomatic benefit from intraarticularly injected superoxide dismutase in patients with osteoarthritis (53), a
disease characterized by very low synovial fluid leukocyte concentrations. The role of oxygen radicals in
arthritis has recently been reviewed (54).
The extent of endothelial cell-crystal interaction in vivo is unknown. That this neglected relationship is a clinically important one is evidenced by the
following observations: 1) There has been an explosion in the knowledge of endothelial physiology and its
regulation of inflammation in the last decade (55,56). 2)
Early pathologic studies demonstrated MSU crystals
in the walls of large blood vessels of patients with gout
(57), as well as in early, noncalcified atherosclerotic
plaques (17). The relationship of clinical gout and
hyperuricemia to atherosclerosis and heart disease
remains an area of active, if controversial, research
(58). 3) Patients with the cholesterol embolization
syndrome typically develop systemic signs such as
fever, weight loss, and a vasculitic rash which, on
biopsy, shows vascular or perivascular inflammatory
infiltrates (12). 4) Dialysis patients frequently develop
vascular calcifications characterized by extensive hydroxyapatite deposits (1 1), intimal hyperplasia, and
FALASCA ET AL
acral gangrene (10). 5 ) Patients with primary or secondary oxalosis are sometimes found to have luminal
as well as medial oxalate deposits (2-5). 6) In vitro
studies have shown that MSU crystals stimulate neutrophil-mediated endothelial damage (59).
In summary, endothelial cells have the capacity
to phagocytize pathogenic crystals and generate superoxide anion in association with this event. Suppression
by cytochalasin B suggests that SOA production by
endothelial cells is at least partially due to crystal
phagocytosis. These phenomena, together with the
previously shown role that endothelial cells may have
in mediating synovial effusions, suggests that endothelial cells may be important in the pathogenesis of acute
crystal-associated inflammation and an important
source of oxygen radicals in many arthritides.
ACKNOWLEDGMENTS
We thank Ann Barnes for her expert secretarial
assistance in the production of the manuscript, and Stephen
N. Mueller, PhD, who performed the Factor VIII and
anti-smooth muscle cell assays, and who provided generous
technical support. We acknowledge the work of Gwen
Harley, who prepared samples for electron microscopy. We
thank the staffs of the Labor and Delivery Suites of Cooper
HospitalLJniversity Medical Center and the West Jersey
Health System for the many umbilical cords they supplied.
REFERENCES
1. Reginato AJ: Calcium oxalate and other crystals or
particles associated with arthritis, Arthritis and Allied
Conditions. Eleventh edition. Edited by DJ McCarty.
Philadelphia, Lea & Febiger, 1989
2. Boquist L, Lindqvist B, Ostberg Y, Steen L: Primary
oxalosis. Am J Med 54:673-681, 1973
3. Arbus GS, Sniderman S: Oxalosis with peripheral gangrene. Arch Pathol97:107-110, 1974
4. Blackburn WE, McRoberts JW, Bhathena D, Vazquez
M, Luke RG: Severe vascular complications in oxalosis
after bilateral nephrectomy. Ann Intern Med 82:44-46,
1975
5 . Op de Hoek CT, Diderich PP, Gratama S, Weijs von
Hofwegen EJ: Oxalosis in chronic renal failure. Proc
Eur Dial Transplant Assoc 17:730-735, 1980
6. Hennigar GF: Drug and chemical injury: environmental
pathology, Anderson’s Pathology. Ninth edition. Edited
by JM Kissane. Philadelphia, CV Mosby, 1990
7. Reginato AJ, Kurnik B: Calcium oxalate and other
crystals associated with kidney disease and arthritis.
Semin Arthritis Rheum 18:198-224, 1989
8. Ross CN, Cassidy MJD, Thompson M, Russell Jones R,
Rees AJ: Proximal cutaneous necrosis associated with
SUPEROXIDE PRODUCTION BY ENDOTHELIAL CELLS
small vessel calcification and renal failure. Q J Med
79:443-450, 1991
9. Reginato AJ, Ferreiro Seoane JL, Barbazan Alvarez C,
Mitja Piferrer J, Vidal Meijon L , Pascual Turon R,
Vasconez F , Rivera ER, Clayburne G, Rothfuss S:
Arthropathy and cutaneous calcinosis in hemodialysis
oxalosis. Arthritis Rheum 29: 1387-1396, 1986
10. Wilkinson SP, Stewart WK, Parham DM, Guthrie W:
Symmetric gangrene of the extremities in late renal
failure: a case report and review of the literature. Q J
Med 67:319-341, 1988
1 1 . Contiguglia SR, Alfrey AC, Miller NL, Runnells DE, Le
Geros RZ: Nature of soft tissue calcification in uraemia.
Kidney Int 4:229-235, 1973
12. Darsee JR: Cholesterol embolization: the great masquerader. South Med J 72:174-180, 1979
13. Stone GC, Wall BA, Oppliger IR, Wener MH, Jolly SL,
Aguirre A, Dwyer R, Simkin PA: A vasculopathy with
deposition of lamba light chain crystals. Ann Intern Med
110~275-278, 1989
14. Edmonds ME, Momson M, Law JW, Watkins PJ:
Medial arterial calcification and diabetic neuropathy . Br
Med J (Clin Res) 284:928-930, 1982
15. Sharp J: Heredo-familial vascular and articular calcifications. Ann Rheum Dis 13:14-27, 1954
16. Carlstrom D, Engfeldt B, Engstrom A, Ringertz N:
Studies on the chemical composition of normal and
abnormal blood vessel walls. 1 . Chemical nature of
calcified vascular deposits. Lab Invest 2:325-335, 1953
17. Platt PN, Malcolm A: Crystals and vessel walls (abstract). Ann Rheum Dis 24 (suppl):109, 1983
18. Simkin PA: Synovial physiology, Arthritis and Allied
Conditions. Eleventh edition. Edited by DJ McCarty.
Philadelphia, Lea & Febiger, 1989
19. Faires JS, McCarty DJ: Acute arthritis in man and dog
after intrasynovial injection of sodium urate crystals.
Lancet II:682-685, 1962
20. Matsubara T, Ziff M: Increased superoxide anion release from human endothelial cells in response to cytokines. J Immunol 137:3295-3298, 1986
21 * Rosen GM, Freeman BA: Detection of superoxide generated by endothelial cells: spin trappindipid peroxidation. Proc Natl Acad Sci U S A 81:7269-7273, 1984
22. Ryan US: The endothelial surface and responses to
injury. Fed Proc 45:lOl-108, 1986
23. Steffan AM, Gendrault JL, McCuskey RS, McCuskey
PA, Kim A: Phagocytosis, an unrecognized property of
murine endothelial liver cells. Hepatology 6:830-836,
1986
24. Hamill RJ, Vann JM, Proctor RA: Phagocytosis of
Staphylococcus aureus by cultured bovine aortic endothelial cells: model for postadherence events in endovascular infections. Infect Immun 54:833-836, 1986
25. Seegmiller JE, Howell RR, Malawista SE: The inflammatory reaction to sodium urate: its possible relation-
115
ship to the genesis of acute gouty arthritis. JAMA
180:469475, 1962
26. Hasselbacher P, Hahn JL: Activation of the alternative
pathway of complement by microcrystalline cholesterol.
Atherosclerosis 373239-245, 1980
27. JaEe EA, Nachman RL, Becker CG, Minick CR: Culture of human endothelial cells derived from umbilical
veins: identification by morphologic and immunologic
criteria. J Clin Invest 52:2745-2756, 1973
28. Maciag T, Cerundolo J, Ilsley S, Kelley PR, Forand R:
An endothelial cell growth factor from bovine hypothalamus: identification and partial characterization. Proc
Natl Acad Sci U S A 765674-5678, 1979
29. Wacker WEC, Ulmer DD. Vallee BL: Metalloenzvmes
and myocardial infarction. 11. Malic acid and lactic
dehydrogenase activities and zinc concentration in serum. N Engl J Med 255M9-456, 1956
30. Schumacher HR, Phelps P: Sequential changes in human polymorphonuclear leukocytes after urate crystal
phagocytosis: an electron microscopic study. Arthritis
Rheum 14513-526, 1971
31. Rook GAW, Steele J, Umar S, Dockrell HM: A simple
method for the solubilisation of reduced NBT, and its
use as a colorimetric assay for activation of human
macrophages by gamma interferon. J Immunol Methods
82:161-167, 1985
32. Simchowitz L, Atkinson JP, Spilberg I: Stimulation of
the respiratory burst in human neutrophils by crystal
phagocytosis. Arthritis Rheum 25: 181-188, 1982
33. Meier B, Radeke HH, Selle S, Raspe HH, Sies H, Resch
K, Habermehl GG: Human fibroblasts release reactive
oxygen species in response to treatment with synovial
fluids from patients suffering from arthritis. Free Radic
Res Commun 8: 149-160, 1989
34. McPhail LC, Henson PM, Johnston RB: Respiratory
burst enzyme in human neutrophils: evidence for multiple mechanisms of activation. J Clin Invest 67:71&716,
1981
35. Goldstein IM, Roos D, Kaplan HB, Weissmann G:
Complement and immunoglobulins stimulate superoxide
production by human leukocytes independently of
phagocytosis. J Clin Invest 5 6 1 155-1 163. 1975
36. Abramson S, Hoffstein ST, Weissmann G: Superoxide
anion generation by human neutrophils exposed to
monosodium urate: effect of protein adsorption and
complement activation. Arthritis Rheum 25: 174-180,
1982
37. Palit J, Di Giovine F, Dickens E , Duff GW, Nuki G:
Extracellular superoxide anion generation in response to
pro-inflammatory crystals: selective stimulation of
monocytes by monosodium urate (abstract). Br J Rheumatol 25 (suppl I):A21, 1986
38. Rosen MS, Baker DG, Schumacher HR Jr, Cherian PV:
Products of polymorphonuclear cell injury inhibit IgG
116
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
enhancement of monosodium urate-induced superoxide
production. Arthritis Rheum 29:1473-1479, 1986
Schumacher HR, Fishbein P, Phelps P, Tse R, Krauser
R: Comparison of sodium urate and calcium pyrophosphate crystal phagocytosis by polymorphonuclear leukocytes: effects of crystal size and other factors. Arthritis Rheum 18 (suppl):783-792, 1975
Di Giovine FS, Malawista SE, Nuki G, Duff GW:
Interleukin 1 (IL 1) as a mediator of crystal arthritis:
stimulation of T cell and synovial fibroblast mitogenesis
by urate crystal-induced IL I . J Immunol 138:3213-3218,
1987
Poubelle PE, de Medicis R, Naccache PH: Monosodium
urate and calcium pyrophosphate crystals differentially
activate the excitation response coupling sequence of
human neutrophils. Biochem Biophys Res Commun
149:649-657, 1987
Terkeltaub RA, Ginsberg MH: The inflammatory reaction to crystals. Rheum Dis Clin North Am 14:353-364,
I988
Brestel EP, McClain EJ, Castranova V: Carrageenan
stimulates reduction of nitroblue tetrazolium by human
neutrophils without membrane depolarization, myeloperoxidase secretion or increased oxygen consumption.
Inflammation 10:425434, 1986
Schumacher HR: The microvasculature of the synovial
membrane of the monkey: ultrastructural studies. Arthritis Rheum 12:387-404, 1969
Schumacher HR: Fate of particulate material arriving at
the synovium via the circulation: an ultrastructural
study. Ann Rheum Dis 32:212-218, 1973
Cotran RS, la Gattuta M, Majno G: Studies on inflammation: fate of intramural vascular deposits induced by
histamine. Am J Pathol 47: 1045-1077, 1965
Garcia JG, Gray LD, Dodson RF, Callahan KS: Asbestos-induced endothelial cell activation and injury: demonstration of fiber phagocytosis and oxidant-dependent
toxicity. Am Rev Respir Dis 138:958-964, 1988
Blake DR, Unsworth J, Outhwaite JM, Morris CJ,
Merry P, Kidd BL, Ballard R, Gray L, Lunec J: Hypoxic-reperfusion injury in the inflamed human joint.
Lancet 1:289-293, 1989
FALASCA ET AL
49. Marcolongo R, Calabria AA, Lalumera M, Gerli R,
Allessandrini C, Cavallo G: The “switch-off’ mechanism of spontaneous resolution of acute gout attack. J
Rheumatol 15: 101-109, 1988
50. Omar HA, Cherry PD, Mortelliti MP, Burke-Wolin T,
Wolin MS: Inhibition of coronary artery superoxide
dismutase attenuates endothelium-dependent and -independent nitrovasodilator relaxation. Circ Res 69501608, 1991
51. Harth M, McCain GA, Orange JF: The effects of auranofin and gold sodium aurothiomalate on the chemiluminescent response of stimulated synovial tissue cells
from patients with rheumatoid arthritis. J Rheumatol
12:881-884, 1985
52. Tiku ML, Liesch JB, Robertson FM: Production of
hydrogen peroxide by rabbit articular chondrocytes:
enhancement by cytokines. J Immunol 145:690-696,
1990
53. McIlwain H, Silverfield JC, Cheatum DE, Poiley J ,
Taborn J, Ignaczak T, Multz CV: Intra-articular orgotein in osteoarthritis of the knee: a placebo-controlled,
efficacy, safety and dosage comparison. Am J Med
87:295-300, 1989
54. Greenwald RA: Oxygen radicals, inflammation, and
arthritis: pathophysiological considerations and implications for treatment. Semin Arthritis Rheum 20:219-240,
1991
55. Fajardo LF: The complexity of endothelial cells: a
review. Am J Clin Pathol 92:241-250, 1989
56. Vane JR, Anggdrd EE, Botting RM: Regulatory functions of the vascular endothelium. N Engl J Med 323:
27-36, 1990
57. Traut EF, Knight AA, Szanto PB, Passerelli EW: Specific vascular changes in gout. JAMA 156591-593, 1954
58. Abbott RD, Brand FN, Kannel WB, Castelli WP: Gout
and coronary artery disease: the Framingham study. J
Clin Epidemiol 41:237-242, 1988
59. Boogaerts MA, Hammerschmidt DE, Roelant C, Verwilghen RL, Jacob HS: Mechanisms of vascular damage
in gout and oxalosis: crystal-induced, granulocyte mediated, endothelial injury. Thromb Haemost 50576-580,
1983
Документ
Категория
Без категории
Просмотров
1
Размер файла
1 174 Кб
Теги
production, crystals, endothelial, phagocytosis, superoxide, anion, culture, cells
1/--страниц
Пожаловаться на содержимое документа