- Email: [email protected]
T cell inhibitor secreted by macrophages and endothelial cells
CLINICAL
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
T Cell Inhibitor
53, 137-150 (1989)
Secreted by Macrophages Endothelial Cells
and
TAKAMICHI KASHIWADO, NANCY OPPENHEIMER-MARKS, AND MORRIS ZIFF Department of Internal Arthritis Research Center,
Medicine (Inflammation Research Unit) The University of Texas Health Science School, Dallas, Texas 75235-9030
and the Harold C. Simmons Center, Southwestern Medical
On stimulation with lipopolysaccharide (LPS), normal human macrophages (M+) and endothelial cells (EC) produced factors which inhibited interleukin 2 (IL-2)-dependent lymphocyte proliferation and PHA plus interleukin 1 (IL-l)-dependent mouse thymocyte proliferation but not IL-l-dependent human libroblast proliferation, suggesting that they were inhibitors of the IL-2 response. In addition, these factors inhibited the production of IL-2 by normal human peripheral blood mononuclear cells (PBMC). The factors also inhibited PBMC proliferation in response to PHA and concanavalin (Con A) but did not inhibit the proliferation of EC, U937 cells, or Epstein-Barr virus-transformed B cells. On Sephadex G200 gel filtration, the inhibitory factors from both M+ and EC were detected almost entirely in a 130- to 150-kDa fraction, but active material was also detected in a IS- to 20-kDa fraction. On isoelectric chromatofocusing of the 130- to 150-kDa fraction, inhibitory activity was associated with fractions eluted at three isoelectric points, pH 7.0, 5.4, and 4.8. The isoelectric fractions isolated from M+ and EC showed similar patterns of inhibition. When 130- to 150-kDa fractions from Sephadex G200 of the M+ and EC supematants were treated with an antibody against a macrophage-derived suppressor factor produced by the human monocytic leukemia cell line THP-1, the activity of both fractions was neutralized. The above findings suggest that normal M+ and EC secrete an identical or closely related inhibitor of IL-2 synthesis and IL-2 response, and that this inhibitor regulates these IL-2-related functions by a suppressive action on the T lymphocyte. 0 1989 Academic Press. lnc
INTRODUCTION
Macrophages (M$) and endothelial cells (EC) play a role in the cellular immune response, not only as antigen-presenting cells, but also as secretors of agents which mediate the immune response (l-8). Both cell types secrete interleukin 1 (IL-l) (2, 7, 8) which stimulates IL-2 production by appropriately activated T lymphocytes (9, 10). The present report describes a suppressor of IL-2 production and response, which is secreted by these cells types. We previously described an inhibitor of the IL-2 response in rheumatoid synovial fluid (11) and noted its cross-reactivity with an antiserum to an IL-2 inhibitor isolated from the human monocytic leukemia cell line THP-1 (12). Previous reports have described IL-l and IL-2 inhibitory factors in the serum of normal human subjects (13, 14), pregnant women (IS) and patients with sarcoidosis (16), acquired immunodeficiency syndrome (17), endotoxin fever (18), retroplacenta (19), and in the urine of febrile patients (20). In vitro production of IL-2 inhibitory factor has been described in cultures of melanoma (21) and glioblastoma cell lines (22). the monocytic cell lines THP-I (23) and U937 (24, 25), monocytes from rheumatoid synovial fluid (26), and monocytes from patients with lepromatous 137 0090-1229189 $ I .50 Copvnghr All nphr\
(‘1 1989 by Academic Pro\. Inc of reprnduct!on in dnv form rewrued
138
KASHIWADO.
OPPENHEIMER-MARKS.
AND
ZIFt
leprosy (27) and scleroderma (28). In the present study, we have investigated factors produced by normal human M+ (adherent cells of blood) and human umbilical vein EC (HUVEC) stimulated with LPS. We have observed that these cells secrete an agent which inhibits (i) IL-2-induced proliferation of the CTLL-2 cell line, (ii) mitogen-stimulated proliferation of peripheral blood mononuclear cells (PBMC), and (iii) production of IL-2 by PHA-stimulated normal PBMC. MATERIALS
AND METHODS
Cell preparation. M$ were separated from normal human PBMC by FicollHypaque gradient centrifugation and suspended in RPM1 1640 containing 15% fetal calf serum (FCS) in plastic culture dishes. After a I-hr incubation, adherent cells were removed by treatment with Puck’s EDTA, washed three times, and suspended in RPM1 containing 3% FCS and antibiotics. More than 95% of these cells were characterized as M$ by nonspecific esterase staining. Human EC were isolated from individual umbilical veins by the method of Jaffe et al. (29). Umbilical veins from individual cords were cannulated, washed with Hanks’ balanced salt solution (HBSS), and Iilled with 0.1% collagenase in HBSS. After a 15-min incubation at room temperature, the detached cells were obtained by washing with HBSS. The cells were centrifuged and resuspended in RPM1 containing 15% FCS, 10% human serum, 25 pg/ml EC growth supplement (ECGS), 50 kg/ml heparin, and antibiotics (complete medium). The cells were cultured overnight in tissue culture flasks at 37°C in humidified 5% CO? and 95% air. The next day, nonadherent cells were removed by extensive washing, and fresh complete medium was added. When the cultures had reached confluency, the flasks were rinsed with Puck’s EDTA and then incubated with Puck’s EDTA and 0.05%’ trypsin to detach the cells. These were then washed, resuspended in complete medium, and seeded into gelatin-coated flasks for further passage. Only third or fourth passage cells were used. Less than 1% of the cells at these passages were stained by a monoclonal antimacrophage antibody (63D3), as determined on a fluorescence-activated cell sorter (FACS). Stimulation ofM4 and EC. M4 and EC were suspended in RPM1 containing 3% FCS at a concentration of 5 x IO5 cells/ml and 1 ml of cells was added to the wells of 24-well, flat-bottomed culture plates (Corning Glass Works, Corning, NY). LPS (final concentration 5 pg/ml) was added to each well and the plates were cultured at 37°C in humidified 5% CO, and 95% air for 24, 48, and 72 hr. The supernatants were then separated, Iiltered through 0.45~km Millipore filters, dialyzed against phosphate-buffered saline (PBS) in dialysis tubing with a cutoff weight of 3.5 kDa, and stored at -70°C before use. As a control, fresh medium (RPM1 containing 3% FCS, 5 kg/ml LPS, and antibiotics) was used. Sephudex G200 gelfilfrution. Crude supernatants were concentrated fivefold on an Amicon filter (cutoff, 10 kDa) and 1 ml of concentrated supernatant was then applied to a column of Sephadex G200 beads in PBS. The column was calibrated with human IgG (150 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), cw-chymotrypsinogen (25 kDa), and ribonuclease (14 kDa). Fractions of 2.5 ml were collected by eluting with PBS and assayed at a I:4 dilution. As a control,
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
139
medium (RPM1 containing 3% FCS and 5 &ml LPS) was dialyzed against PBS, concentrated fivefold, and 1 ml was applied to a column. DEAE 52 cellulose ion-exchange chromatography. Two milliliters of the 130- to 150-kDa fractions obtained by Sephadex G200 gel filtration was dialyzed against 20 mM Tris buffer, pH 7.4, applied to a 1.2 x 8-cm column of DEAE 52 cellulose and eluted with increasing concentrations of NaCl varying from 0 to 1.0 M. Two milliliters of each fraction was collected, dialyzed against PBS, and assayed at 1:4 dilution. IL-2-dependent cell proliferation assay. Fifty microliters of test samples of inhibitor was added to individual wells of 96-well flat-bottomed microtiter plates. To each well were then added 4 x lo3 CTLL-2 cells, a murine IL-Zdependent cell line, suspended in RPM1 1640, containing 10% FCS, 50 ~.LM 2-mercaptoethanol, and antibiotics. They were cultured in a final volume of 200 ~1 with or without exogenous recombinant IL-2 (Biogen, Geneva, Switzerland) for 48 hr, pulsed for the last 8 hr with 1 &i/well of [3H]TdR, and harvested. To determine the optimum quantity of IL-2 to be added in this assay, we established that the CTLL-2 cell response to IL-2 was linear between 2.5 and 20 U/ml. In most experiments, 10 U/ml of IL-2 was, therefore, used. Results were expressed as (i) cpm, (ii) percentage control response, or (iii) percentage inhibition, as follows: % Control response = cpm test sample/cpm control % Inhibition = 100 - % control response.
X 100
Znhibition of IL-2 synthesis. Normal PBMC were isolated by Ficoll-Hypaque centrifugation and incubated with inhibitory or control fractions at a final dilution of 1:4 in RPM1 1640 at a final cell concentration of 1 x lo6 cells/ml for 1,4,8, and 24 hr at 37°C. The cells were washed three times and 1 x lo6 of such viable PBMC in 1 ml were then cultured for 24 hr with 1 pg/rnl of PHA in RPM1 1640 containing 2% FCS at 37°C. The culture supernatants were then obtained by centrifugation and assayed for IL-2 activity in the CTLL-2 proliferation assay. Mouse thymocyte proliferation IL-l assay. C3H/HeJ mouse thymocytes (1 x 106) were added to individual wells of 96-well microtiter plates in RPM1 containing 5% FCS and 1 l&ml PHA. Fifty microliters of test sample was then added (final volume 200 l.i.1)and the mixtures were cultured for 72 hr in the presence or absence of 2 U/ml of rIL-1 (Biogen). The cultures were pulsed with 1 &i/well of [3H]TdR during the last 18 hr and harvested. Fibroblust proliferation assay. Fibroblasts from normal human infant foreskin in their 4th to 10th passage were added in the amount of 1 x lo4 cells/well to 96-well flat-bottomed microtiter plates in a volume of 150 p,l together with 50 l~,l of inhibitory fraction in RPM1 1640 containing 5% FCS and cultured for 72 hr in the presence or absence of 2 U/ml of rIL-1. Cultures were pulsed for the last 18 hr with 1 &i/well of [3H]TdR. The medium was removed and the cells were harvested, after a IO-min treatment with 0.1 mM EDTA and 0.125% trypsin, to measure [3H]TdR uptake. Preparation of anti-IgG serum against a macrophage-derived immunosuppressor factor (MDSF). A rabbit was immunized with a preparation of a MDSF iso-
140
KASHIWADO.
OPPENHEIMER-MARKS.
AND
ZIFF
lated from a supernatant of the human monocytic leukemia cell line, THP-I , by Dr. T. Krakauer, National Institutes of Health (Bethesda, MD) (23), which inhibits both IL-2 production and the response to IL-2. To obtain the supernatant, I x IO6 cells/ml of THP- 1 cells were cultured with 50 pg/ml of silica in RPM1 1640 for 72 hr, the supernatant was fractionated on AcA 54 Ultrogel, and the 67-kDa fraction containing MDSF activity was obtained. Rabbits were then immunized with this fraction as previously described (12) and the IgG fractions of the immune sera isolated. As control, IgG from a nonimmunized rabbit was obtained in a similar manner. Effect of inhibitor on spontaneous proliferation of EC, U937 cell line, and EBV-transformed BlI cell line. EC (I X 104) in the third or fourth passage were seeded into the individual wells of gelatin-coated, flat-bottomed, 96-well microtiter plates along with 50 ~1 of the 130- to 150-kDa fraction from Sephadex G200 in a final dilution of 1:4 in RPM1 containing 5% FCS (final volume of 200 p.l), cultured for 48 hr, and pulsed with 1 kCi/well of 13H]TdR during the last 18 hr before harvesting. Cells (2 x 104) of the U937 monocytic cell line were seeded into individual wells along with the same inhibitor samples in RPM1 containing 10% FCS and antibiotics, cultured for 5 days, and pulsed with 1 kCi/well of [3H]TdR during the last 18 hr before harvesting. Cells (2 x 104) of the EBV-transformed B cell line B 11, which proliferates spontaneously, were added to individual wells along with the same inhibitor samples in RPM1 containing 10% FCS, cultured for 72 hr, and pulsed during the last 18 hr with 1 l..&i/well of [-7H]TdR before harvesting. Proliferation of PHA and Con A-stimulated normal human PBMC. Normal human PBMC were added to the wells of 96-well, flat-bottomed microtiter plates along with the 130- to 1.50-kDa inhibitory fraction in 1:4 dilution at a final concentration of 5 x 10’ cells/ml in RPM1 1640 containing 10% FCS. Then, PHA (1 pg/ml) or Con A (10 pg/ml) was added and the cells were cultured for 72 hr and pulsed for the last 18 hr with 1 &i/well of [3H]TdR before harvesting. Chromatofocusing of gelfiltration fractions. The 130- to 150-kDa fractions from a Sephadex G200 column were pooled, dialyzed overnight at 4°C against 0.025 M imidazole-HCl buffer. and submitted to isoelectric chromatofocusing in the pH range of 7.4-4.0. A 1 x 15-cm bed of PBE 94 gel (Pharmacia, Uppsala, Sweden) was equilibrated at 4°C with 0.02 M imidazole start buffer, pH 7.4. The dialyzed sample was applied to the column, and a gradient of pH from 7.4 to 4.0 was established by elution at a flow rate of 30 ml/hr with polybuffer 74 (Pharmacia) adjusted to pH 4.0 with 0.1 N HCI. Fractions of 2 ml were collected and their pH measured. These were dialyzed overnight at 4°C against PBS and stored at - 70°C until use. RESULTS Effects of supernatants of LPS-stimulated Md, and EC on CTLL-2, thymocyte, andfibroblast proliferations. Supernatants of M$ and EC stimulated by LPS for 48 hr significantly depressed IL-Zstimulated CTLL-2 proliferation (P < 0.001) (Fig. IA). They also significantly reduced IL-l-stimulated thymocyte proliferation
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
CTLL-2 A‘0 X
.-. o-0
MB
Supernotant ,I
.-. n--a
EC
Supernatont I,
Thymocyte .-. MB supernotant t, 0-o
30
E
a 200 .-z ‘0 z g to-2
E
Wl,h IL-2
+=i \ \
T CELL
20 S-1
cTT z .-. A--o
i I
EC
141
INHIBITOR
Fibroblast ,
Supsrnatant ,I
1-z
MB SuPeyatant
.-.
EC
Supsrnotant
Y -r -1
With
IL-l
I’I
I
Without
IL-2
IIt
B--cx===K==s _ 0 24 48 72 Hours of LPS Stlmulatlon
Wlthout
11
0 Hours of LPS Stimulation
IL-l
11
24 48 72
Hours of LPS Stimulation
FIG. 1. Effects of supematants of LPS-stimulated M+ and EC on CTLL-2, thymocyte, and fibroblast proliferation. Each of four supematants, obtained by stimulating suspensions of M+ and EC with LPS for increasing periods of time, were assayed for their effects on CTLL-2, thymocyte, and fibroblast proliferation in triplicate. Supematants were assayed on (A) CTLL-2 proliferation at 1:4 dilution. with or without exogenous IL-2, 10 U/ml (background proliferation, 643 2 216 cpm); (B) thymocyte proliferation, with or without exogenous IL-l, 2 U/ml (background proliferation with PHA alone, 2768 & 349 cpm); and (C) fibroblast proliferation, with or without exogenous IL-I, 2 U/ml (background proliferation, 4986 2 1253 cpm). Values represent the mean of six experiments.
(P < O.Ol), though in the absence of IL-l, both mildly enhanced thymocyte proliferation (Fig. 1B). Finally, neither supernatant suppressed IL-l-induced fibroblast proliferation and in the absence of IL-l, both variably enhanced fibroblast proliferation (Fig. IC). Cell viabilities of the stimulated M+ were 93.0 2 3.5% (0 hr), 91.9 ? 1.3% (24 hr), 90.9 + 2.9% (48 hr), and 88.9 ? 2.5% (72 hr) and those of the EC were 87.4 2 1.2% (0 hr), 87.0 ? 0.9% (24 hr), 85.5 rfr 0.8% (48 hr), and 83.2 ? 3.5 % (72 hr). Changes in viability with time were not statistically significant (P < 0.1). Neither supernatant exerted a cytotoxic effect on CTLL-2 cells, thymocytes, or fibroblasts either in the presence or in the absence of complement. These results indicate that M+ and EC supernatants stimulated with LPS for 48 hr contain one or more inhibitors of the response to IL-2 of CTLL-2 cells and the response to IL-l of PHA-stimulated thymocytes but not of the ILl-stimulated proliferation of fibroblasts. Effects of Sephadex G200 fractions of M$ and EC supernatants on CTLL-2 proliferation. When the supernatants of LPS-stimulated M$ and EC were frac-
tionated on Sephadex G200, both M4 and EC supernatant fractions in the molecular weight range of 100-160 kDa inhibited IL-2-stimulated CTLL-2 proliferation. Maximum activity was observed in the 130- to 150-kDa range. Fractions of 15-20 kDa mol wt also inhibited weakly (Fig. 2). Effects of Sephadex G200 fractions of M+ and EC supernatants on thymocyte proliferation. When the Sephadex G200 fractions were assayed in the IL-
l-induced
thymocyte
proliferation
assay, the 130- to I50-kDa
and 15- to 20-kDa
142
KASHIWADO,
OPPENHEIMER-MARKS,
AND
O-O
Medbum
.-. A-A
EC
S~pernotont I, Wlthout
‘1
Lliqg@F~8~~=*~=*=~&oA
v-l
1 20
25
ZIFF
30 35 40 45 Froctlon Number
50
IL-2
60
FIG. 2. Effects of Sephadex G200 fractions of MI+ and EC supematants on CTLL-2 proliferation. M4, and EC supematants were fractionated on Sephadex G200 and assayed at 1:4 dilution in triplicate for their effects on CTLL-2 proliferation with or without exogenous IL-2 (10 U/ml). Assays were done on each of three fractionated supematants of M+ and four fractionated supematants of EC. The patterns in each case were similar to that shown (Background proliferation, 647 ? 145 cpm).
fractions of both supernatants inhibited thymocyte proliferation. The 50- to 60kDa and 16-kDa fractions enhanced proliferation in the absence of exogenous IL-I, presumably because they contained IL-l (Fig. 3). Effects of Sephadex G200 fractions of M+ and EC superantants onfibroblast proliferation. When the effects of the Sephadex G200 fractions on fibroblast pro-
X E e .-6 z
-f-’ 2o-7
z E i
lo-
.-•
.-• 0-o
MI
O-0
Medium
.-. A-A
EC
Supsrnotont I,
supsrnatoIlt I, Without
IL-1
Froctlon Number 3. Effects of Sephadex G200 fractions of M+ and EC supematants on thymocyte proliferation. The same fractions examined in Fig. 2 were assayed in the PHA-stimulated thymocyte proliferation assay with or without exogenous IL-I (2 U/ml) (background proliferation with PHA alone, 1946 * 316 cpm). FIG.
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
143
liferation were examined, no inhibition of IL-l-stimulated proliferation was observed, but as with thymocyte proliferation, the 50- to 60-kDa and 16-kDa fractions enhanced proliferation in the absence of exogenous IL-l (Fig. 4). When the concentration of exogenous IL-l was decreased, the effects of these fractions on fibroblast proliferation were unchanged (data not shown). Effects of IL-2 concentration and CTLL-2 cell concentration on the activity of the inhibitorfraction. The inhibitory activity of the 130- to 150-kDa fraction from both M+ and EC supernatants diminished as the CTLL-2 concentration was increased and disappeared almost completely when the cell concentration was increased fivefold (Table 1). On the other hand, the inhibitory activity was only partially and not significantly reduced when the IL-2 concentration was raised from 10 to 80 U/ml (Table 2). The greater reduction of the inhibitory activity with an increase in cell concentration rather than with an increase in IL-2 concentration suggests that the action of this inhibitor is on the CTLL-2 cell and not on the action of IL-2. When 1 ml of an active, 1:4 diluted inhibitory fraction from Sephadex G200 was absorbed twice with lo6 CTLL-2 cells, all inhibitory activity was removed, indicating that the inhibitor was, in fact, bound to the cells. Effect of inhibitor fractions on IL-2 production. The 130- to 150-kDa inhibitor fractions were preincubated with normal PBMC for increasing periods of time and the treated PBMC were subsequently stimulated with PHA. Following a 24-hr preincubation period, the washed cells were stimulated with 1 &ml of PHA for 24 hr, and the IL-2 content of the culture supematants was assayed. It was observed that the IL-2 activities of the supematants of the cells preincubated with inhibitor fractions were significantly decreased and that the percentage reduction increased with the period of preincubation. Preincubation with either inhibitor
0 4> H” 2 0 t t t .-. MI supernatant II
m4 x
E =:m 0 UC t t
0-o
20 i
A-A A-A
EC supsrnotont I, With
IL-1
WIthout
frI $ WI
1 20
25
30
35
Fraction
40
45
50
IL-l
60
Number
FIG. 4. Effects of Sephadex G200 fractions on fibroblast proliferation. The same supematant fractions as utilized in Figs. 1 and 2 were assayed on human fibroblast proliferation with or without exogenous IL-1 (2 U/ml) (background proliferation, 4275 -+ 523 cpm).
144
KASHIWADO,
OPPENHEIMER-MARKS,
AND ZIFF
TABLE I EFFECTS OF CTLL-2 CELL CONCENTRATION ON IL-~-INDUCED CTLL-2 PROLIFERATION IN THY PRESENCEOF 130- ro 150-kDa INHIBITOR FRACTIONS” Source of 130- to ISO-kDa fraction
2 x IO4
WJ % Inhibition EC % Inhibition Medium control
7.057 -+ 603h (54.7) 8,485 t 224’ (45.6) 15,584 2 1498
CTLL-2 cells/ml 5 x IO4
1 x IO5
cpm CTLL-2 cells 32,213 ? 1831h 93.358 2 3259’ (31.8) (2.3) 37,527 2 1451” 92,842 -t- 6315’ (19.4) (3.5) 46,554 ? 2490 96,256 k 5114
2x IO 61,327 I 2479’ (1.1) 64,596 2 1299’ (-4.2) 62,017 2 2010
0 Three preparations of inhibitory fraction were assayed on IL-2-induced CTLL-2 proliferation at 1:4 dilution, each in triplicate. Values represent mean -t- SD of three experiments. Background cpm without IL-2 were 576 5 132 (2 X 104). 663 t 190 (5 x IO?, 826 t 221 (I x lo’), and 774 * 295 (2 A IO’). Concentration of IL-2 was IO U/ml and values in parentheses represent percentage inhibition of control values. P values represent significance of the percentage of inhibition of the control values exerted by both M$ and EC supernatants. h P < 0.001. ’ Not significant.
fraction for less than 8 hr resulted in only a minimal reduction in IL-2 production. However, preincubation with the M$ inhibitor fraction for 24 hr led to a 46% reduction (P < 0.001) and a similar preincubation with the EC inhibitor fraction led to a 30% reduction (P < 0.001) (Fig. 5). Similar patterns of inhibition of IL-2 production were observed using two additional 130- to 150-kDa fractions from each of two M$ and two EC supernatants. In these experiments, the thorough washing of the inhibitor-treated PBMC prior to incubation with PHA and the subsequent fourfold dilution of the PHA supernatants in the assay of IL-2 activity TABLE 2 EFFECTS OF IL-2 CONCENTRATION ON IL-~-INDUCED CTLL-2 PROLIFERATION IN THE PRESENCE OF THE 130- TO 150-kDa INHIBII-OR FRACTIONS ~~ .~~ ~~-. ~~..~~~~~ ~~~ ~~~~~~~ ~~~ Source of IL-2 U/ml 130- to ISO-kDa fraction IO 20 40 80 m+ % Inhibition EC % Inhibition Medium control
7.057 t 603’ (54.7) 8,485 t 224’ (45.6) 15,584 lr 1498
cpm CTLL-2 cells 17,157 2 1675’ 18,701 2 860b (30.0) (29.0) 18.294 2 1012’ 19.024 2 1352’ (25.8) (27.7) 24,642 2 1588 26,304 k 1267
13,731 t 739b (32.6) 16,583 -c 1664b (18.7) 20,387 2 953
u Three preparations of the l30- to 150-kDa inhibitory fraction were assayed on IL-2-induced CTLL2 proliferation at 1:4 dilution each in triplicate. Values represent mean 2 SD of three experiments. Background cpm in the absence of IL-2 were 576 2 132. Concentration of CTLL-2 cells was 2 x 104/ml and values in parentheses represent percentage inhibition of control values. P values represent significance of the percentage of inhibition of the control values by both M$ and EC supematants. bP < 0.001.
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
145
g 100 c g.2
9-0 A+
ES t aJN
0;
M0 EC
Inhibitory inhibitory
Froctlon Fraction
50-
g$H
3-I .c &AS _
L.LI/
0
I
I
4 8 Duration of Preincubation
24 (hrs)
FIG. 5. Effects of inhibitor fractions on IL-2 production. The 130- to 150-kDa inhibitor fractions from M+ and EC supematants were preincubated with 1 x IO6 PBMC/ml at I:4 dilution for 1,4,8, and 24 hr and the cells were then washed three times. Viable PBMC (1 x 106/ml), which had been so treated with either one of the inhibitor fractions or with control fractions from the column, were cultured with 1 pg/ml of PHA for 24 hr in RPM1 1640 containing 2% FCS. The supematants thus obtained were assayed at 1:4 dilution in quintuplicate for their activity in stimulating CTLL-2 proliferation. Results are expressed as the percentage of decrease in IL-2 activity. Counts per minute of the IL-2 activities of PBMC supematants preincubated with the control fraction were 19847 +- 1687 (0 hr), 16745 2 2132 (1 hr). 16324 2 954 (4 hr). 15473 jI 1521 (8 hr). and 27235 ? 994 (24 hr).
ruled out the possibility synthesis observed.
of carryover of inhibitor
as a factor in the decrease in IL-2
Effect of inhibitorfractions on spontaneous proliferation of EC, U937 cells, and the EBV-transformed Bll cell line. When the 130- to I50-kDa inhibitor fractions
were examined for their capacity to inhibit the spontaneous proliferation of EC, U937 cells, and the EBV-transformed B cell line, B 11, no effect on the spontaneous proliferation of all three cell types was observed (data not shown). Effect of an antibody to macrophage-derived suppressor factor on the inhibitory activity of the 130- to 150-kDa fractions. A preparation of IgG from a rabbit
which had been immunized with a MDSF secreted by the THP-1 cell line on the inhibition of IL-2-stimulated CTLL-2 proliferation was examined. The IgG antibody preparation reduced the inhibition produced by the 130- to 150-kDa fractions of the M+ and EC supematants as well as that of the MDSF preparation itself, while IgG from a nonimmunized rabbit had no effect (Fig. 6). These results suggest that the inhibitory factors secreted by M+, EC, and THP-1 cells share similar epitopes which may be related to their inhibitory activities. Inhibitory activity of a fraction obtained by DEAE 52 ion-exchange chromatography. The 130- to 150-kDa fractions of M$ and EC supernatants were applied
to a DEAE 52 cellulose column in 20 mM Tris buffer, pH 7.4, and eluted with NaCl. The single peak obtained which eluted in an identical manner from both supernatants was dialyzed against PBS and tested for IL-2 inhibitory activity. Such activity was noted in the eluted material from both supematants (data not shown). The eluted material also inhibited the normal PBMC response to PHA (P < 0.001) and Con A (P < 0.001) (Table 3). The activities of the M+ and EC
146
KASHIWADO.
OPPENHEIMER-MARKS, Anlt-MDSF
IgG
AND .-.
I.40
4-A
EC Inhibitory MDSF
Fractton
M0 lnhlbdory EC lnhlbitory MDSF
Fraction Fraction
l -m Control
0-o
IgGA-A
O-O
I
lnhlbltory
I
I
I
I
0
1
10
100
Antibody
ZIFF
Concentration
Fracflon
I
1000
(pg/ml)
FIG. 6. Effects of antibody against macrophage-derived suppressor factor (MDSF) on activity of the 130- to 150~kDa inhibitor fractions. The activity of 1:4 dilutions of the 130- to 150-kDa fractions in inhibiting IL-2-induced CTLL-2 proliferation was measured in triplicate in the presence of increasing concentrations of an antibody against a MDSF isolated from the supematant of the human monocytic leukemia cell line THP-1. Results are expressed as percentage control proliferation. The cpm of control proliferation were 17,436 ? 1321, cpm of proliferation with M+ inhibitory fraction were 7183 2 598 (41.2%), with EC inhibitory fraction, 9185 ? 698 (52.6%), and with the IL-2 inhibitory fraction from THP-1 (MDSF), 4313 2 276 (24.7%). Similar patterns were obtained using two additional preparations of Mb and EC inhibitor fractions.
preparations obtained by DEAE 52 chromatography were both resistant to heating at 56°C for 30 min (data not shown). Effect of chromatofocused inhibitor fractions on IL-2-induced CTLL-2 cell proliferation. When the 130- to SO-kDa fractions of M+ and EC supernatants were submitted to isoelectric chromatofocusing, the inhibitory activity directed against IL-Zinduced CTLL-2 cell proliferation was eluted in an identical manner from TABLE 3 EFFECTSOFIL-~INHIBITORY FRACTION ONPHA ANDCONA-STIMULATED PBMC PROLIFERATIONS -. cpm PBMC Source of inhibitor
PHA
Con A
Wb ECb PBS control
30,487 2 27%’ 34,509 f 3327’ 41,492 ” 2168
8,238 ” 447’ 9,850 t 725’ 15,794 2 1940
a Three preparations of Sephadex G200, 130- to 150-kDa fractions of both M+ and EC supematants were faltered through a DEAE 52 cellulose column and assayed for their effect on PBMC proliferation in response to PHA (1 t&nl) and Con A (10 &nl) at 1:4 dilution. Values represent mean 2 SD of three experiments. Cell concentration of the PBMC was 5 x 105/ml and background cpm were 1524 t 361 in the absence of mitogen. b Void volume from DEAE 52 cellulose column. = P < 0.001.
MACROPHAGE
PH
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
147
100 9;k
--
7.0
-0- -0, ‘o “0
6.0
‘\ a-o.
.-.
MQ Supernatant
Ah
EC Supernatant
\
5.0 4.0 6
14 18 22 26 30 34 Fraction Number FIG. 7. Effect of isoelectric chromatofocused fractions on IL-2 response. Fractions (130-150 kDa) from Sephadex G200 were combined and purified by chromatofocusing in the pH range of 7.4-4.0. Alternate fractions were assayed for their effect on CTLL-2 proliferation in the presence of IL-2 (10 U/ml) at 1:4 dilution in triplicate. Results are expressed as mean percentage inhibition of the control value. Background proliferation was 547 rf- 276 cpm without IL-2 and 16478 -t 1275 cpm with IL-2.
both the M+ and EC supematant 5.4, and 4.8 (Fig. 7).
10
in three peaks with pZ values in each case of 7.0, DISCUSSION
We have demonstrated inhibition by supematants of LPS-stimulated normal human M$ and EC of both the synthesis of IL-2 in cultures of mitogen-stimulated, normal human PBMC and the response to IL-2 of the murine IL-Zdependent CTLL-2 cell line. On Sephadex G200 gel filtration, the inhibitory activity for both the synthesis of IL-2 and the response to IL-2 was detected in 130- to 150-kDa and 15- to 20-kDa fractions. Most of the inhibitory activity was in the high molecular weight fraction. The high molecular weight inhibitory agents from M+ and EC were similar in size, isoelectric chromatofocusing pattern, and in their action on CTLL-2 cells, IL- l-stimulated thymocytes and IL- 1-stimulated fibroblasts . Moreover, both preparations were neutralized in a similar fashion by a rabbit antibody against MDSF, isolated from the human monocytic leukemia cell line, THP-1. These findings strongly suggested that the inhibitors from M4, and EC were closely related or identical. Specifically, both agents inhibited PHA-stimulated IL-2 production by PBMC, IL-Zinduced CTLL-2 cell proliferation, IL-l-induced PHA-stimulated thymocyte proliferation, and PHA and Con A-stimulated normal human PBMC proliferation. They had no effect on IL-l-induced fibroblast proliferation or the spontaneous proliferation of human umbilical vein EC, the monocytic U937 cell line, or the EBV-transformed B cell line, B 11. The action of the inhibitor, in the case of IL-2 production, appeared to be exerted on the T cell since the IL-Zproducing PBMC were preincubated with the inhibitor and washed prior to addition of PHA, and since the degree of inhibition increased with the duration of the preincubation period. In the inhibition of the response to IL-2, the action appeared to be exerted on the CTLL-2 cell because the inhibitory effect disap-
148
KASHIWADO,
OPPENHEIMER-MARKS,
AND
ZIFF
peared when the concentration of the CTLL-2 cells was increased while it decreased only partially on increasing the concentration of exogenous IL-2. Finally. absorption of the Sephadex G200 inhibitory fraction with CTLL-2 cells completely removed the activity, indicating that the inhibitor was bound to the cells. Although IL-I promotes the production of IL-2 (30), it is unlikely that the observed inhibition of IL-2-dependent cellular proliferation was due to an IL-1 inhibitor since the present agent inhibited IL-2-induced CTLL-2 proliferation, a phenomenon in which IL-I plays no role. As to the observed inhibition of thymocyte proliferation by the active agent, this response is dependent on both IL- 1 and IL-2 (9, IO), and it is likely, therefore, that the effect on thymocyte proliferation resulted from inhibition of the IL-2 phase of this assay (3 1, 32). Lotz el ul. (33) described an inhibitor of IL-I in supernatants of monocytes separated from rheumatoid peripheral blood mononuclear cells which down-regulated the ILl-dependent synthesis of IL-2 by normal peripheral blood mononuclear cells. Culture supernatants of rheumatoid synovial fluid mononuclear cells also had a pronounced inhibitory action on IL-l-stimulated thymocyte proliferation (26). In our experiments, the 130- to 150-kDa fraction of supernatants of LPS-stimulated monocytes and EC suppressed not only IL-l-induced thymocyte proliferation, but also the response to IL-2 of CTLL-2 cells and its production by PBMC. The possibility that the inhibition of IL-2 synthesis was due to an IL-l inhibitor of the type described by Lotz et al. is rendered unlikely by the fact that our supernatants did not affect IL-l-stimulated tibroblast proliferation (34, 35). The apparent presence of an IL-2 inhibitor in our preparations and not in the supernatants of Lotz et al. may be due to the additional stimulation provided by LPS in our cultures. Production of an inhibitor of IL-2 synthesis has been described in cultures of adherent cells from lepromatous leprosy patients (27). the macrophage-like cell line U937 (24, 25), and a melanoma cell line (21). In addition, secretion of an inhibitor of the IL-2 response by glioblastoma cells (22) has also been reported. Krakauer has described a MDSF obtained from the human monocytic leukemia cell line, THP-1, which inhibited both the production of IL-2 and its effects on target cells (23). The molecular weight of this factor, eluted from AcA 54 Ultrogel, was 60-70 kDa in contrast to the molecular weight of the 130-150 kDa of our inhibitor eluted from Sephadex G200. Despite this difference in size, the factor described by us cross-reacted with an antibody produced against the MDSF from THP-I, indicating a close relationship between these two agents. Antigen-presenting cells appear to play the additional role of being immunoregulatory cells, as indicated by the capacity of cells of the monocyte-macrophage series to secrete a number of immunosuppressive agents, i.e., PGE, (36.37), IL-l inhibitor (25-27), and the present T cell inhibitor which diminishes IL-2 synthesis and IL-2 action. In view of the fact that EC have also been demonstrated to act as antigen-presenting cells (6, 38, 39) and recent suggestions that antigen presentation may take place on the endothelial surface (40-42), our observation of the secretion of an inhibitor of T cell function by human umbilical vein EC is of special interest. It suggests that modulation of the immune response by an agent secreted by the endothelium may be initiated on the endothelial surface or in the immediate perivascular area.
MACROPHAGE
AND ENDOTHELIAL
CELL-DERIVED
T CELL INHIBITOR
149
ACKNOWLEDGMENTS We thank Dr. Hugo Jasin for his help in the production of IL-2 inhibitor antibody, Dr. Theresa Krakauer for making available a preparation of MDSF from the THP-I cell line, Dr. Druie Cavender for providing us with the cell lines U937 and Bl 1, and Mr. Dale Edelbaum and Mrs. Carla Suson for their excellent assistance in these experiments. This work was supported by NIH Program/Project Grant AMO9989.
REFERENCES 1. Unanue, E. R., The regulatory role of macrophages in antigenic stimulation. Part 2: Symbolic relationship between lymphocytes and macrophages. Adv. Immunol. 1, 136, 1981. 2. Beller, D. I., and Unanue, E. R., Reciprocal regulation of macrophage and T cell function by way of soluble mediators. Lymphokines 6, 25, 1981. 3. Inaba, K., and Steinman, R. M., Accessory cell-T lymphocyte interactions: Antigen-dependent and -independent clustering. 1. Exp. Med. 163, 247, 1986. 4. Nathan, C. F., Secretory products of macrophages. J. C/in. Invest. 79, 319, 1987. 5. Barr, R. S., Goldsmith, J. C., and Owen, W. G.. Properties of endothelial cells in culture: Recent studies. Lymphokines 7, 249, 1982. 6. Hirschberg, H., Bergh, 0. J., and Thorsby, E., Antigen presenting properties of human vascular endothelial cells. J. Exp. Med. 152, 249s 1980. 7. Stem. D. M., Bank, I., Naworth, P. P., Cassimeris, J., Kisie, W., Fenton, W., II, Dinarello, C., Chess, L., and Jaffe, E. A., Self-regulation of procoagulant events on the endothelial cell surface. J. Exp. Med. 162, 1223, 1985. 8. Miossec, P.. Cavender, D., and Ziff, M., Production of interleukin 1 by human endothelial cells, J. Immunol. 136, 2486, 1986. 9. Smith, K. A., Lachman, L. B., Oppenheim, J. J.. and Favata. M. F., The functional relationship of the interleukins. .I. Exp. Med. 151, 1551, 1980. 10. Oppenheim, J. J., and Gery. I., Interleukin 1 is more than interleukin. Immunol. Today 3, 113. 1982. Il. Miossec, P., Kashiwado, T., and Ziff. M., Inhibitor of interleukin 2 in rheumatoid synovial fluid. Arthritis Rheum. 30, 121, 1987. 12. Kashiwado, T.. Miossec, P., Oppenheimer-Marks, N., and Ziff, M., Interleukin 2 inhibitor in rheumatoid synovial fluid. Arthritis Rheum. 30, 1339, 1987. 13. Wolf, R. L., and Andreoni, J., Soluble inhibitory factor (SIF) in normal human serum. Cell. Immunol. 61, 299, 1982. 14. Djeu, J. Y., Kasahara, T., Balow, J. E., and Toskos, G. C., Decreased interleukin 2 inhibitor in sera of patients with autoimmune disorders. Clin. Exp. Immunol. 65, 279, 1986. 15. Nicholas, N. S.. Panayi, G. S., and Touri, A. M., Human pregnancy serum inhibit interleukin 2 production. Clin. Exp. Immunol. 58, 587, 1984. 16. Nath. I., Ellner, J. J., Bouknight, R., Tomford, J. W.. Kleinhenz, M. E., and Edmons. K. L., Interrelationship of immunoregulatory cells and serum factors in sarcoidosis. J. Immunol. 125, 1071, 1980. 17. Siegel, J. P.. Djeu, J. Y., Stocks, N. I., Masur, H., Glemann, E. P., and Quinnan, G. V., Jr., Sera from patients with acquired immunodeficiency syndrome inhibit production interleukin-2 by normal lymphocytes. J. C/in. Invest. 75, 1957, 1985. 18. Dinarello, C. A., Rosenwasser, L. J., and Wolff, S. M., Demonstration of a circulating suppressor factor of thymocyte proliferation during endotoxin fever in human. J. Zmmunol. 127, 2517, 1981. 19. Nicholas, N. S., and Panayi, G. S., Inhibition of IL-2 production by retroplacental sera: A possible mechanism for human fetal allograft survival. Amer. J. Reprod. Immunol. Microbial. 9, 6, 1985. 20. Liao, Z., Grimshaw. R. S., and Rosenstreich, R. S., Identification of a specific interleukin 1 inhibitor in the urine of febrile patients. J. Exp. Med. 159, 126, 1984. 21. Hersey, P., Bindon, C., Czemiecki, M., Spurling, A., Wass, J., and McCarthy, W. H., Inhibition of interleukin 2 production by factors released from tumor cells. J. Immunol. 131, 2837, 1983.
150
KASHIWADO.
OPPENHEIMER-MARKS,
AND
ZIFF
22. Fontana, A., Hengartner, H., Tribolet, N., and Weber, E.. Glioblastoma cells release interleukin 1 and factors inhibiting interleukin 2 mediated effects. J. Immunol. 132, 1837, 1984. 23. Krakauer, T., A macrophage derived factor that inhibits the production and action of interleukin 2. J. Leukocyte Biol. 38, 429, 1985. 24. Fujiwara, H., and Elhter, J. J., Spontaneous production of a suppressor factor by the human macrophage-like cell line U937.1. Suppression of interleukin 1, interleukin 2 and mitogen induced blastogenesis in mouse thymocytes. J. Immunol. 136, 181, 1986. 25. Fujiwara, H., Toossi, Z., Ohnishi, K., Edmonds, K., and Ellner, J. J., Spontaneous production of a suppressor factor by a human macrophage-like cell line U937. II. Suppression of antigen- and mitogen-induced blastogenesis. IL-2 production and IL-2 receptor expression in T lymphocytes. J. Immunol.
138,
197,
1987.
26. Lotz, M., Tsoukas, C. D., Robinson, C. A., Dinarello, C. A., Carson, D. A., and Vaughan, J. H., Basis for defective responses of rheumatoid arthritis synovial fluid lymphocytes to anti-CD3(T3) antibodies. J. Clin. Invest. 78, 713, 1986. 27. Nath, I., Jayaraman, J., Sathis, M., Bhutani, L. K., and Sharama, A. K., Inhibition of interleukin 2 production by adherent cell factors from lepromatous leprosy patients. Clin. Exp. Zmmunol. 58, 531, 1984. 28. Sandborg, C. I., Berman, M. A.. Andrews, B. S., Mirick, G. R., and Friou, G. J.. Increased production of an interleukin 1 (IL-l) inhibitor with Bbroblast stimulating activity by mononuclear cells from patients with scleroderma. Clin. Exp. Immunol. 66, 312, 1986. 29. Jaffe, E. A., Nachman, R. L., Becker, C. G., and Minick, C. R., Culture of human endothelial cells derived from umbilical veins. J. Clin. Invest. 52, 2745, 1973. 30. Smith, K. A., Gilbride, K. J., and Favata, M. F., Lymphocyte activating factor promotes T-cell growth factor production by cloned murine lymphoma cells. Nature (London) 287, 853, 1980. 3 1. Oppenheim, J. J., Stadler, B. M., Siraganian, R. P., Mage, M., and Mathieson, B., Lymphokines: Their role in lymphocyte responses. Properties of interleukin 1. Fed. Proc. 41, 257, 1982. 32. Caplan, B , and Rothenberg, E., High-level secretion of interleukin 2 by a subset of proliferating thymic lymphoblasts. J. Immunol. 133, 1983, 1984. 33. Lotz, M., Tsoukas, C. D., Fong, S., Dinarello, C. A., Carson, D. A., and Vaughan, J. H., Release of lymphokines after infection with Epstein-Barr virus in vitro. II. A monocyte-dependent inhibitor of interleukin 1 downregulates the production of interleukin 2 and interferon-y in rheumatoid arthritis. J. Immunol. 136, 3643, 1986. 34. Postlethwaite, A. E., Lachman, L. B., and Kang, A. H., Induction of fibroblast proliferation by interleukin 1 derived from human monocytic leukemia cells. Arthritis Rheum. 27, 995, 1984. 35. Kom, J. H.. Torres, D., and Downie. D., Clonal heterogeneity in the fibroblast response to mononuclear cell derived mediators. Arthritis Rheum. 27, 174, 1984. 36. Goodwin, J. S., Bankhurst, A. D., and Morley, J., Suppression of human T-cell mitogenesis by prostaglandin. J. Exp. Med. 146, 1719, 1977. 37. Farrar, W. L., and Humes, J. L.. The role of arachiodonic acid metabolism in the activities of interleukin 1 and 2. J. Immunol. 135, 1153, 1985. 38. Ashida, E., Johnson, A., and Lipsky, P. E., Human endothelial cell-lymphocyte interactions. Endothelial cells function as accessory cells necessary for mitogen-induced T lymphocyte activation in vitro. J. Clin. Invest. 671, 1490, 1981. 39. Nunez, G., Ball, E. J., and Stastny, P., Accessory cell function of human endothehal cells, I. A subpopulation of Ia positive cells is required for antigen presentation. J. Zmmunol. 131, 666, 1983. 40. Pober, J. S., Gimbrone, M. A., Jr., Collins, T., Cotran, R. S., Ault, K. A., Fiers, W., Krensky, A. M., Clayberger, C., Reiss, C. S., and Burakoff, S. J., Interactions of T lymphocytes with human vascular endothelial cell surface antigens. Immunobiology 168, 483, 1984. 41. Wagner, C. R., Vetto, R. M., and Burger, D. R., Expression of I-region-associated antigen (Ia) and interleukin 1 by subcultured human endothelial cells. Cell. Zmmunol. 93, 91, 1985. 42. Geppert, T. D., and Lipsky, P. E., Antigen presentation by interferon-y-treated endothelial cells and fibroblasts: Differential ability to function as antigen-presenting cells despite comparable Ia expression. J. Immunol. 135, 3750. 1985. Received February 27, 1989; accepted May 16, 1989
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
T Cell Inhibitor
53, 137-150 (1989)
Secreted by Macrophages Endothelial Cells
and
TAKAMICHI KASHIWADO, NANCY OPPENHEIMER-MARKS, AND MORRIS ZIFF Department of Internal Arthritis Research Center,
Medicine (Inflammation Research Unit) The University of Texas Health Science School, Dallas, Texas 75235-9030
and the Harold C. Simmons Center, Southwestern Medical
On stimulation with lipopolysaccharide (LPS), normal human macrophages (M+) and endothelial cells (EC) produced factors which inhibited interleukin 2 (IL-2)-dependent lymphocyte proliferation and PHA plus interleukin 1 (IL-l)-dependent mouse thymocyte proliferation but not IL-l-dependent human libroblast proliferation, suggesting that they were inhibitors of the IL-2 response. In addition, these factors inhibited the production of IL-2 by normal human peripheral blood mononuclear cells (PBMC). The factors also inhibited PBMC proliferation in response to PHA and concanavalin (Con A) but did not inhibit the proliferation of EC, U937 cells, or Epstein-Barr virus-transformed B cells. On Sephadex G200 gel filtration, the inhibitory factors from both M+ and EC were detected almost entirely in a 130- to 150-kDa fraction, but active material was also detected in a IS- to 20-kDa fraction. On isoelectric chromatofocusing of the 130- to 150-kDa fraction, inhibitory activity was associated with fractions eluted at three isoelectric points, pH 7.0, 5.4, and 4.8. The isoelectric fractions isolated from M+ and EC showed similar patterns of inhibition. When 130- to 150-kDa fractions from Sephadex G200 of the M+ and EC supematants were treated with an antibody against a macrophage-derived suppressor factor produced by the human monocytic leukemia cell line THP-1, the activity of both fractions was neutralized. The above findings suggest that normal M+ and EC secrete an identical or closely related inhibitor of IL-2 synthesis and IL-2 response, and that this inhibitor regulates these IL-2-related functions by a suppressive action on the T lymphocyte. 0 1989 Academic Press. lnc
INTRODUCTION
Macrophages (M$) and endothelial cells (EC) play a role in the cellular immune response, not only as antigen-presenting cells, but also as secretors of agents which mediate the immune response (l-8). Both cell types secrete interleukin 1 (IL-l) (2, 7, 8) which stimulates IL-2 production by appropriately activated T lymphocytes (9, 10). The present report describes a suppressor of IL-2 production and response, which is secreted by these cells types. We previously described an inhibitor of the IL-2 response in rheumatoid synovial fluid (11) and noted its cross-reactivity with an antiserum to an IL-2 inhibitor isolated from the human monocytic leukemia cell line THP-1 (12). Previous reports have described IL-l and IL-2 inhibitory factors in the serum of normal human subjects (13, 14), pregnant women (IS) and patients with sarcoidosis (16), acquired immunodeficiency syndrome (17), endotoxin fever (18), retroplacenta (19), and in the urine of febrile patients (20). In vitro production of IL-2 inhibitory factor has been described in cultures of melanoma (21) and glioblastoma cell lines (22). the monocytic cell lines THP-I (23) and U937 (24, 25), monocytes from rheumatoid synovial fluid (26), and monocytes from patients with lepromatous 137 0090-1229189 $ I .50 Copvnghr All nphr\
(‘1 1989 by Academic Pro\. Inc of reprnduct!on in dnv form rewrued
138
KASHIWADO.
OPPENHEIMER-MARKS.
AND
ZIFt
leprosy (27) and scleroderma (28). In the present study, we have investigated factors produced by normal human M+ (adherent cells of blood) and human umbilical vein EC (HUVEC) stimulated with LPS. We have observed that these cells secrete an agent which inhibits (i) IL-2-induced proliferation of the CTLL-2 cell line, (ii) mitogen-stimulated proliferation of peripheral blood mononuclear cells (PBMC), and (iii) production of IL-2 by PHA-stimulated normal PBMC. MATERIALS
AND METHODS
Cell preparation. M$ were separated from normal human PBMC by FicollHypaque gradient centrifugation and suspended in RPM1 1640 containing 15% fetal calf serum (FCS) in plastic culture dishes. After a I-hr incubation, adherent cells were removed by treatment with Puck’s EDTA, washed three times, and suspended in RPM1 containing 3% FCS and antibiotics. More than 95% of these cells were characterized as M$ by nonspecific esterase staining. Human EC were isolated from individual umbilical veins by the method of Jaffe et al. (29). Umbilical veins from individual cords were cannulated, washed with Hanks’ balanced salt solution (HBSS), and Iilled with 0.1% collagenase in HBSS. After a 15-min incubation at room temperature, the detached cells were obtained by washing with HBSS. The cells were centrifuged and resuspended in RPM1 containing 15% FCS, 10% human serum, 25 pg/ml EC growth supplement (ECGS), 50 kg/ml heparin, and antibiotics (complete medium). The cells were cultured overnight in tissue culture flasks at 37°C in humidified 5% CO? and 95% air. The next day, nonadherent cells were removed by extensive washing, and fresh complete medium was added. When the cultures had reached confluency, the flasks were rinsed with Puck’s EDTA and then incubated with Puck’s EDTA and 0.05%’ trypsin to detach the cells. These were then washed, resuspended in complete medium, and seeded into gelatin-coated flasks for further passage. Only third or fourth passage cells were used. Less than 1% of the cells at these passages were stained by a monoclonal antimacrophage antibody (63D3), as determined on a fluorescence-activated cell sorter (FACS). Stimulation ofM4 and EC. M4 and EC were suspended in RPM1 containing 3% FCS at a concentration of 5 x IO5 cells/ml and 1 ml of cells was added to the wells of 24-well, flat-bottomed culture plates (Corning Glass Works, Corning, NY). LPS (final concentration 5 pg/ml) was added to each well and the plates were cultured at 37°C in humidified 5% CO, and 95% air for 24, 48, and 72 hr. The supernatants were then separated, Iiltered through 0.45~km Millipore filters, dialyzed against phosphate-buffered saline (PBS) in dialysis tubing with a cutoff weight of 3.5 kDa, and stored at -70°C before use. As a control, fresh medium (RPM1 containing 3% FCS, 5 kg/ml LPS, and antibiotics) was used. Sephudex G200 gelfilfrution. Crude supernatants were concentrated fivefold on an Amicon filter (cutoff, 10 kDa) and 1 ml of concentrated supernatant was then applied to a column of Sephadex G200 beads in PBS. The column was calibrated with human IgG (150 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), cw-chymotrypsinogen (25 kDa), and ribonuclease (14 kDa). Fractions of 2.5 ml were collected by eluting with PBS and assayed at a I:4 dilution. As a control,
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
139
medium (RPM1 containing 3% FCS and 5 &ml LPS) was dialyzed against PBS, concentrated fivefold, and 1 ml was applied to a column. DEAE 52 cellulose ion-exchange chromatography. Two milliliters of the 130- to 150-kDa fractions obtained by Sephadex G200 gel filtration was dialyzed against 20 mM Tris buffer, pH 7.4, applied to a 1.2 x 8-cm column of DEAE 52 cellulose and eluted with increasing concentrations of NaCl varying from 0 to 1.0 M. Two milliliters of each fraction was collected, dialyzed against PBS, and assayed at 1:4 dilution. IL-2-dependent cell proliferation assay. Fifty microliters of test samples of inhibitor was added to individual wells of 96-well flat-bottomed microtiter plates. To each well were then added 4 x lo3 CTLL-2 cells, a murine IL-Zdependent cell line, suspended in RPM1 1640, containing 10% FCS, 50 ~.LM 2-mercaptoethanol, and antibiotics. They were cultured in a final volume of 200 ~1 with or without exogenous recombinant IL-2 (Biogen, Geneva, Switzerland) for 48 hr, pulsed for the last 8 hr with 1 &i/well of [3H]TdR, and harvested. To determine the optimum quantity of IL-2 to be added in this assay, we established that the CTLL-2 cell response to IL-2 was linear between 2.5 and 20 U/ml. In most experiments, 10 U/ml of IL-2 was, therefore, used. Results were expressed as (i) cpm, (ii) percentage control response, or (iii) percentage inhibition, as follows: % Control response = cpm test sample/cpm control % Inhibition = 100 - % control response.
X 100
Znhibition of IL-2 synthesis. Normal PBMC were isolated by Ficoll-Hypaque centrifugation and incubated with inhibitory or control fractions at a final dilution of 1:4 in RPM1 1640 at a final cell concentration of 1 x lo6 cells/ml for 1,4,8, and 24 hr at 37°C. The cells were washed three times and 1 x lo6 of such viable PBMC in 1 ml were then cultured for 24 hr with 1 pg/rnl of PHA in RPM1 1640 containing 2% FCS at 37°C. The culture supernatants were then obtained by centrifugation and assayed for IL-2 activity in the CTLL-2 proliferation assay. Mouse thymocyte proliferation IL-l assay. C3H/HeJ mouse thymocytes (1 x 106) were added to individual wells of 96-well microtiter plates in RPM1 containing 5% FCS and 1 l&ml PHA. Fifty microliters of test sample was then added (final volume 200 l.i.1)and the mixtures were cultured for 72 hr in the presence or absence of 2 U/ml of rIL-1 (Biogen). The cultures were pulsed with 1 &i/well of [3H]TdR during the last 18 hr and harvested. Fibroblust proliferation assay. Fibroblasts from normal human infant foreskin in their 4th to 10th passage were added in the amount of 1 x lo4 cells/well to 96-well flat-bottomed microtiter plates in a volume of 150 p,l together with 50 l~,l of inhibitory fraction in RPM1 1640 containing 5% FCS and cultured for 72 hr in the presence or absence of 2 U/ml of rIL-1. Cultures were pulsed for the last 18 hr with 1 &i/well of [3H]TdR. The medium was removed and the cells were harvested, after a IO-min treatment with 0.1 mM EDTA and 0.125% trypsin, to measure [3H]TdR uptake. Preparation of anti-IgG serum against a macrophage-derived immunosuppressor factor (MDSF). A rabbit was immunized with a preparation of a MDSF iso-
140
KASHIWADO.
OPPENHEIMER-MARKS.
AND
ZIFF
lated from a supernatant of the human monocytic leukemia cell line, THP-I , by Dr. T. Krakauer, National Institutes of Health (Bethesda, MD) (23), which inhibits both IL-2 production and the response to IL-2. To obtain the supernatant, I x IO6 cells/ml of THP- 1 cells were cultured with 50 pg/ml of silica in RPM1 1640 for 72 hr, the supernatant was fractionated on AcA 54 Ultrogel, and the 67-kDa fraction containing MDSF activity was obtained. Rabbits were then immunized with this fraction as previously described (12) and the IgG fractions of the immune sera isolated. As control, IgG from a nonimmunized rabbit was obtained in a similar manner. Effect of inhibitor on spontaneous proliferation of EC, U937 cell line, and EBV-transformed BlI cell line. EC (I X 104) in the third or fourth passage were seeded into the individual wells of gelatin-coated, flat-bottomed, 96-well microtiter plates along with 50 ~1 of the 130- to 150-kDa fraction from Sephadex G200 in a final dilution of 1:4 in RPM1 containing 5% FCS (final volume of 200 p.l), cultured for 48 hr, and pulsed with 1 kCi/well of 13H]TdR during the last 18 hr before harvesting. Cells (2 x 104) of the U937 monocytic cell line were seeded into individual wells along with the same inhibitor samples in RPM1 containing 10% FCS and antibiotics, cultured for 5 days, and pulsed with 1 kCi/well of [3H]TdR during the last 18 hr before harvesting. Cells (2 x 104) of the EBV-transformed B cell line B 11, which proliferates spontaneously, were added to individual wells along with the same inhibitor samples in RPM1 containing 10% FCS, cultured for 72 hr, and pulsed during the last 18 hr with 1 l..&i/well of [-7H]TdR before harvesting. Proliferation of PHA and Con A-stimulated normal human PBMC. Normal human PBMC were added to the wells of 96-well, flat-bottomed microtiter plates along with the 130- to 1.50-kDa inhibitory fraction in 1:4 dilution at a final concentration of 5 x 10’ cells/ml in RPM1 1640 containing 10% FCS. Then, PHA (1 pg/ml) or Con A (10 pg/ml) was added and the cells were cultured for 72 hr and pulsed for the last 18 hr with 1 &i/well of [3H]TdR before harvesting. Chromatofocusing of gelfiltration fractions. The 130- to 150-kDa fractions from a Sephadex G200 column were pooled, dialyzed overnight at 4°C against 0.025 M imidazole-HCl buffer. and submitted to isoelectric chromatofocusing in the pH range of 7.4-4.0. A 1 x 15-cm bed of PBE 94 gel (Pharmacia, Uppsala, Sweden) was equilibrated at 4°C with 0.02 M imidazole start buffer, pH 7.4. The dialyzed sample was applied to the column, and a gradient of pH from 7.4 to 4.0 was established by elution at a flow rate of 30 ml/hr with polybuffer 74 (Pharmacia) adjusted to pH 4.0 with 0.1 N HCI. Fractions of 2 ml were collected and their pH measured. These were dialyzed overnight at 4°C against PBS and stored at - 70°C until use. RESULTS Effects of supernatants of LPS-stimulated Md, and EC on CTLL-2, thymocyte, andfibroblast proliferations. Supernatants of M$ and EC stimulated by LPS for 48 hr significantly depressed IL-Zstimulated CTLL-2 proliferation (P < 0.001) (Fig. IA). They also significantly reduced IL-l-stimulated thymocyte proliferation
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
CTLL-2 A‘0 X
.-. o-0
MB
Supernotant ,I
.-. n--a
EC
Supernatont I,
Thymocyte .-. MB supernotant t, 0-o
30
E
a 200 .-z ‘0 z g to-2
E
Wl,h IL-2
+=i \ \
T CELL
20 S-1
cTT z .-. A--o
i I
EC
141
INHIBITOR
Fibroblast ,
Supsrnatant ,I
1-z
MB SuPeyatant
.-.
EC
Supsrnotant
Y -r -1
With
IL-l
I’I
I
Without
IL-2
IIt
B--cx===K==s _ 0 24 48 72 Hours of LPS Stlmulatlon
Wlthout
11
0 Hours of LPS Stimulation
IL-l
11
24 48 72
Hours of LPS Stimulation
FIG. 1. Effects of supematants of LPS-stimulated M+ and EC on CTLL-2, thymocyte, and fibroblast proliferation. Each of four supematants, obtained by stimulating suspensions of M+ and EC with LPS for increasing periods of time, were assayed for their effects on CTLL-2, thymocyte, and fibroblast proliferation in triplicate. Supematants were assayed on (A) CTLL-2 proliferation at 1:4 dilution. with or without exogenous IL-2, 10 U/ml (background proliferation, 643 2 216 cpm); (B) thymocyte proliferation, with or without exogenous IL-l, 2 U/ml (background proliferation with PHA alone, 2768 & 349 cpm); and (C) fibroblast proliferation, with or without exogenous IL-I, 2 U/ml (background proliferation, 4986 2 1253 cpm). Values represent the mean of six experiments.
(P < O.Ol), though in the absence of IL-l, both mildly enhanced thymocyte proliferation (Fig. 1B). Finally, neither supernatant suppressed IL-l-induced fibroblast proliferation and in the absence of IL-l, both variably enhanced fibroblast proliferation (Fig. IC). Cell viabilities of the stimulated M+ were 93.0 2 3.5% (0 hr), 91.9 ? 1.3% (24 hr), 90.9 + 2.9% (48 hr), and 88.9 ? 2.5% (72 hr) and those of the EC were 87.4 2 1.2% (0 hr), 87.0 ? 0.9% (24 hr), 85.5 rfr 0.8% (48 hr), and 83.2 ? 3.5 % (72 hr). Changes in viability with time were not statistically significant (P < 0.1). Neither supernatant exerted a cytotoxic effect on CTLL-2 cells, thymocytes, or fibroblasts either in the presence or in the absence of complement. These results indicate that M+ and EC supernatants stimulated with LPS for 48 hr contain one or more inhibitors of the response to IL-2 of CTLL-2 cells and the response to IL-l of PHA-stimulated thymocytes but not of the ILl-stimulated proliferation of fibroblasts. Effects of Sephadex G200 fractions of M$ and EC supernatants on CTLL-2 proliferation. When the supernatants of LPS-stimulated M$ and EC were frac-
tionated on Sephadex G200, both M4 and EC supernatant fractions in the molecular weight range of 100-160 kDa inhibited IL-2-stimulated CTLL-2 proliferation. Maximum activity was observed in the 130- to 150-kDa range. Fractions of 15-20 kDa mol wt also inhibited weakly (Fig. 2). Effects of Sephadex G200 fractions of M+ and EC supernatants on thymocyte proliferation. When the Sephadex G200 fractions were assayed in the IL-
l-induced
thymocyte
proliferation
assay, the 130- to I50-kDa
and 15- to 20-kDa
142
KASHIWADO,
OPPENHEIMER-MARKS,
AND
O-O
Medbum
.-. A-A
EC
S~pernotont I, Wlthout
‘1
Lliqg@F~8~~=*~=*=~&oA
v-l
1 20
25
ZIFF
30 35 40 45 Froctlon Number
50
IL-2
60
FIG. 2. Effects of Sephadex G200 fractions of MI+ and EC supematants on CTLL-2 proliferation. M4, and EC supematants were fractionated on Sephadex G200 and assayed at 1:4 dilution in triplicate for their effects on CTLL-2 proliferation with or without exogenous IL-2 (10 U/ml). Assays were done on each of three fractionated supematants of M+ and four fractionated supematants of EC. The patterns in each case were similar to that shown (Background proliferation, 647 ? 145 cpm).
fractions of both supernatants inhibited thymocyte proliferation. The 50- to 60kDa and 16-kDa fractions enhanced proliferation in the absence of exogenous IL-I, presumably because they contained IL-l (Fig. 3). Effects of Sephadex G200 fractions of M+ and EC superantants onfibroblast proliferation. When the effects of the Sephadex G200 fractions on fibroblast pro-
X E e .-6 z
-f-’ 2o-7
z E i
lo-
.-•
.-• 0-o
MI
O-0
Medium
.-. A-A
EC
Supsrnotont I,
supsrnatoIlt I, Without
IL-1
Froctlon Number 3. Effects of Sephadex G200 fractions of M+ and EC supematants on thymocyte proliferation. The same fractions examined in Fig. 2 were assayed in the PHA-stimulated thymocyte proliferation assay with or without exogenous IL-I (2 U/ml) (background proliferation with PHA alone, 1946 * 316 cpm). FIG.
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
143
liferation were examined, no inhibition of IL-l-stimulated proliferation was observed, but as with thymocyte proliferation, the 50- to 60-kDa and 16-kDa fractions enhanced proliferation in the absence of exogenous IL-l (Fig. 4). When the concentration of exogenous IL-l was decreased, the effects of these fractions on fibroblast proliferation were unchanged (data not shown). Effects of IL-2 concentration and CTLL-2 cell concentration on the activity of the inhibitorfraction. The inhibitory activity of the 130- to 150-kDa fraction from both M+ and EC supernatants diminished as the CTLL-2 concentration was increased and disappeared almost completely when the cell concentration was increased fivefold (Table 1). On the other hand, the inhibitory activity was only partially and not significantly reduced when the IL-2 concentration was raised from 10 to 80 U/ml (Table 2). The greater reduction of the inhibitory activity with an increase in cell concentration rather than with an increase in IL-2 concentration suggests that the action of this inhibitor is on the CTLL-2 cell and not on the action of IL-2. When 1 ml of an active, 1:4 diluted inhibitory fraction from Sephadex G200 was absorbed twice with lo6 CTLL-2 cells, all inhibitory activity was removed, indicating that the inhibitor was, in fact, bound to the cells. Effect of inhibitor fractions on IL-2 production. The 130- to 150-kDa inhibitor fractions were preincubated with normal PBMC for increasing periods of time and the treated PBMC were subsequently stimulated with PHA. Following a 24-hr preincubation period, the washed cells were stimulated with 1 &ml of PHA for 24 hr, and the IL-2 content of the culture supematants was assayed. It was observed that the IL-2 activities of the supematants of the cells preincubated with inhibitor fractions were significantly decreased and that the percentage reduction increased with the period of preincubation. Preincubation with either inhibitor
0 4> H” 2 0 t t t .-. MI supernatant II
m4 x
E =:m 0 UC t t
0-o
20 i
A-A A-A
EC supsrnotont I, With
IL-1
WIthout
frI $ WI
1 20
25
30
35
Fraction
40
45
50
IL-l
60
Number
FIG. 4. Effects of Sephadex G200 fractions on fibroblast proliferation. The same supematant fractions as utilized in Figs. 1 and 2 were assayed on human fibroblast proliferation with or without exogenous IL-1 (2 U/ml) (background proliferation, 4275 -+ 523 cpm).
144
KASHIWADO,
OPPENHEIMER-MARKS,
AND ZIFF
TABLE I EFFECTS OF CTLL-2 CELL CONCENTRATION ON IL-~-INDUCED CTLL-2 PROLIFERATION IN THY PRESENCEOF 130- ro 150-kDa INHIBITOR FRACTIONS” Source of 130- to ISO-kDa fraction
2 x IO4
WJ % Inhibition EC % Inhibition Medium control
7.057 -+ 603h (54.7) 8,485 t 224’ (45.6) 15,584 2 1498
CTLL-2 cells/ml 5 x IO4
1 x IO5
cpm CTLL-2 cells 32,213 ? 1831h 93.358 2 3259’ (31.8) (2.3) 37,527 2 1451” 92,842 -t- 6315’ (19.4) (3.5) 46,554 ? 2490 96,256 k 5114
2x IO 61,327 I 2479’ (1.1) 64,596 2 1299’ (-4.2) 62,017 2 2010
0 Three preparations of inhibitory fraction were assayed on IL-2-induced CTLL-2 proliferation at 1:4 dilution, each in triplicate. Values represent mean -t- SD of three experiments. Background cpm without IL-2 were 576 5 132 (2 X 104). 663 t 190 (5 x IO?, 826 t 221 (I x lo’), and 774 * 295 (2 A IO’). Concentration of IL-2 was IO U/ml and values in parentheses represent percentage inhibition of control values. P values represent significance of the percentage of inhibition of the control values exerted by both M$ and EC supernatants. h P < 0.001. ’ Not significant.
fraction for less than 8 hr resulted in only a minimal reduction in IL-2 production. However, preincubation with the M$ inhibitor fraction for 24 hr led to a 46% reduction (P < 0.001) and a similar preincubation with the EC inhibitor fraction led to a 30% reduction (P < 0.001) (Fig. 5). Similar patterns of inhibition of IL-2 production were observed using two additional 130- to 150-kDa fractions from each of two M$ and two EC supernatants. In these experiments, the thorough washing of the inhibitor-treated PBMC prior to incubation with PHA and the subsequent fourfold dilution of the PHA supernatants in the assay of IL-2 activity TABLE 2 EFFECTS OF IL-2 CONCENTRATION ON IL-~-INDUCED CTLL-2 PROLIFERATION IN THE PRESENCE OF THE 130- TO 150-kDa INHIBII-OR FRACTIONS ~~ .~~ ~~-. ~~..~~~~~ ~~~ ~~~~~~~ ~~~ Source of IL-2 U/ml 130- to ISO-kDa fraction IO 20 40 80 m+ % Inhibition EC % Inhibition Medium control
7.057 t 603’ (54.7) 8,485 t 224’ (45.6) 15,584 lr 1498
cpm CTLL-2 cells 17,157 2 1675’ 18,701 2 860b (30.0) (29.0) 18.294 2 1012’ 19.024 2 1352’ (25.8) (27.7) 24,642 2 1588 26,304 k 1267
13,731 t 739b (32.6) 16,583 -c 1664b (18.7) 20,387 2 953
u Three preparations of the l30- to 150-kDa inhibitory fraction were assayed on IL-2-induced CTLL2 proliferation at 1:4 dilution each in triplicate. Values represent mean 2 SD of three experiments. Background cpm in the absence of IL-2 were 576 2 132. Concentration of CTLL-2 cells was 2 x 104/ml and values in parentheses represent percentage inhibition of control values. P values represent significance of the percentage of inhibition of the control values by both M$ and EC supematants. bP < 0.001.
MACROPHAGE
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
145
g 100 c g.2
9-0 A+
ES t aJN
0;
M0 EC
Inhibitory inhibitory
Froctlon Fraction
50-
g$H
3-I .c &AS _
L.LI/
0
I
I
4 8 Duration of Preincubation
24 (hrs)
FIG. 5. Effects of inhibitor fractions on IL-2 production. The 130- to 150-kDa inhibitor fractions from M+ and EC supematants were preincubated with 1 x IO6 PBMC/ml at I:4 dilution for 1,4,8, and 24 hr and the cells were then washed three times. Viable PBMC (1 x 106/ml), which had been so treated with either one of the inhibitor fractions or with control fractions from the column, were cultured with 1 pg/ml of PHA for 24 hr in RPM1 1640 containing 2% FCS. The supematants thus obtained were assayed at 1:4 dilution in quintuplicate for their activity in stimulating CTLL-2 proliferation. Results are expressed as the percentage of decrease in IL-2 activity. Counts per minute of the IL-2 activities of PBMC supematants preincubated with the control fraction were 19847 +- 1687 (0 hr), 16745 2 2132 (1 hr). 16324 2 954 (4 hr). 15473 jI 1521 (8 hr). and 27235 ? 994 (24 hr).
ruled out the possibility synthesis observed.
of carryover of inhibitor
as a factor in the decrease in IL-2
Effect of inhibitorfractions on spontaneous proliferation of EC, U937 cells, and the EBV-transformed Bll cell line. When the 130- to I50-kDa inhibitor fractions
were examined for their capacity to inhibit the spontaneous proliferation of EC, U937 cells, and the EBV-transformed B cell line, B 11, no effect on the spontaneous proliferation of all three cell types was observed (data not shown). Effect of an antibody to macrophage-derived suppressor factor on the inhibitory activity of the 130- to 150-kDa fractions. A preparation of IgG from a rabbit
which had been immunized with a MDSF secreted by the THP-1 cell line on the inhibition of IL-2-stimulated CTLL-2 proliferation was examined. The IgG antibody preparation reduced the inhibition produced by the 130- to 150-kDa fractions of the M+ and EC supematants as well as that of the MDSF preparation itself, while IgG from a nonimmunized rabbit had no effect (Fig. 6). These results suggest that the inhibitory factors secreted by M+, EC, and THP-1 cells share similar epitopes which may be related to their inhibitory activities. Inhibitory activity of a fraction obtained by DEAE 52 ion-exchange chromatography. The 130- to 150-kDa fractions of M$ and EC supernatants were applied
to a DEAE 52 cellulose column in 20 mM Tris buffer, pH 7.4, and eluted with NaCl. The single peak obtained which eluted in an identical manner from both supernatants was dialyzed against PBS and tested for IL-2 inhibitory activity. Such activity was noted in the eluted material from both supematants (data not shown). The eluted material also inhibited the normal PBMC response to PHA (P < 0.001) and Con A (P < 0.001) (Table 3). The activities of the M+ and EC
146
KASHIWADO.
OPPENHEIMER-MARKS, Anlt-MDSF
IgG
AND .-.
I.40
4-A
EC Inhibitory MDSF
Fractton
M0 lnhlbdory EC lnhlbitory MDSF
Fraction Fraction
l -m Control
0-o
IgGA-A
O-O
I
lnhlbltory
I
I
I
I
0
1
10
100
Antibody
ZIFF
Concentration
Fracflon
I
1000
(pg/ml)
FIG. 6. Effects of antibody against macrophage-derived suppressor factor (MDSF) on activity of the 130- to 150~kDa inhibitor fractions. The activity of 1:4 dilutions of the 130- to 150-kDa fractions in inhibiting IL-2-induced CTLL-2 proliferation was measured in triplicate in the presence of increasing concentrations of an antibody against a MDSF isolated from the supematant of the human monocytic leukemia cell line THP-1. Results are expressed as percentage control proliferation. The cpm of control proliferation were 17,436 ? 1321, cpm of proliferation with M+ inhibitory fraction were 7183 2 598 (41.2%), with EC inhibitory fraction, 9185 ? 698 (52.6%), and with the IL-2 inhibitory fraction from THP-1 (MDSF), 4313 2 276 (24.7%). Similar patterns were obtained using two additional preparations of Mb and EC inhibitor fractions.
preparations obtained by DEAE 52 chromatography were both resistant to heating at 56°C for 30 min (data not shown). Effect of chromatofocused inhibitor fractions on IL-2-induced CTLL-2 cell proliferation. When the 130- to SO-kDa fractions of M+ and EC supernatants were submitted to isoelectric chromatofocusing, the inhibitory activity directed against IL-Zinduced CTLL-2 cell proliferation was eluted in an identical manner from TABLE 3 EFFECTSOFIL-~INHIBITORY FRACTION ONPHA ANDCONA-STIMULATED PBMC PROLIFERATIONS -. cpm PBMC Source of inhibitor
PHA
Con A
Wb ECb PBS control
30,487 2 27%’ 34,509 f 3327’ 41,492 ” 2168
8,238 ” 447’ 9,850 t 725’ 15,794 2 1940
a Three preparations of Sephadex G200, 130- to 150-kDa fractions of both M+ and EC supematants were faltered through a DEAE 52 cellulose column and assayed for their effect on PBMC proliferation in response to PHA (1 t&nl) and Con A (10 &nl) at 1:4 dilution. Values represent mean 2 SD of three experiments. Cell concentration of the PBMC was 5 x 105/ml and background cpm were 1524 t 361 in the absence of mitogen. b Void volume from DEAE 52 cellulose column. = P < 0.001.
MACROPHAGE
PH
AND
ENDOTHELIAL
CELL-DERIVED
T CELL
INHIBITOR
147
100 9;k
--
7.0
-0- -0, ‘o “0
6.0
‘\ a-o.
.-.
MQ Supernatant
Ah
EC Supernatant
\
5.0 4.0 6
14 18 22 26 30 34 Fraction Number FIG. 7. Effect of isoelectric chromatofocused fractions on IL-2 response. Fractions (130-150 kDa) from Sephadex G200 were combined and purified by chromatofocusing in the pH range of 7.4-4.0. Alternate fractions were assayed for their effect on CTLL-2 proliferation in the presence of IL-2 (10 U/ml) at 1:4 dilution in triplicate. Results are expressed as mean percentage inhibition of the control value. Background proliferation was 547 rf- 276 cpm without IL-2 and 16478 -t 1275 cpm with IL-2.
both the M+ and EC supematant 5.4, and 4.8 (Fig. 7).
10
in three peaks with pZ values in each case of 7.0, DISCUSSION
We have demonstrated inhibition by supematants of LPS-stimulated normal human M$ and EC of both the synthesis of IL-2 in cultures of mitogen-stimulated, normal human PBMC and the response to IL-2 of the murine IL-Zdependent CTLL-2 cell line. On Sephadex G200 gel filtration, the inhibitory activity for both the synthesis of IL-2 and the response to IL-2 was detected in 130- to 150-kDa and 15- to 20-kDa fractions. Most of the inhibitory activity was in the high molecular weight fraction. The high molecular weight inhibitory agents from M+ and EC were similar in size, isoelectric chromatofocusing pattern, and in their action on CTLL-2 cells, IL- l-stimulated thymocytes and IL- 1-stimulated fibroblasts . Moreover, both preparations were neutralized in a similar fashion by a rabbit antibody against MDSF, isolated from the human monocytic leukemia cell line, THP-1. These findings strongly suggested that the inhibitors from M4, and EC were closely related or identical. Specifically, both agents inhibited PHA-stimulated IL-2 production by PBMC, IL-Zinduced CTLL-2 cell proliferation, IL-l-induced PHA-stimulated thymocyte proliferation, and PHA and Con A-stimulated normal human PBMC proliferation. They had no effect on IL-l-induced fibroblast proliferation or the spontaneous proliferation of human umbilical vein EC, the monocytic U937 cell line, or the EBV-transformed B cell line, B 11. The action of the inhibitor, in the case of IL-2 production, appeared to be exerted on the T cell since the IL-Zproducing PBMC were preincubated with the inhibitor and washed prior to addition of PHA, and since the degree of inhibition increased with the duration of the preincubation period. In the inhibition of the response to IL-2, the action appeared to be exerted on the CTLL-2 cell because the inhibitory effect disap-
148
KASHIWADO,
OPPENHEIMER-MARKS,
AND
ZIFF
peared when the concentration of the CTLL-2 cells was increased while it decreased only partially on increasing the concentration of exogenous IL-2. Finally. absorption of the Sephadex G200 inhibitory fraction with CTLL-2 cells completely removed the activity, indicating that the inhibitor was bound to the cells. Although IL-I promotes the production of IL-2 (30), it is unlikely that the observed inhibition of IL-2-dependent cellular proliferation was due to an IL-1 inhibitor since the present agent inhibited IL-2-induced CTLL-2 proliferation, a phenomenon in which IL-I plays no role. As to the observed inhibition of thymocyte proliferation by the active agent, this response is dependent on both IL- 1 and IL-2 (9, IO), and it is likely, therefore, that the effect on thymocyte proliferation resulted from inhibition of the IL-2 phase of this assay (3 1, 32). Lotz el ul. (33) described an inhibitor of IL-I in supernatants of monocytes separated from rheumatoid peripheral blood mononuclear cells which down-regulated the ILl-dependent synthesis of IL-2 by normal peripheral blood mononuclear cells. Culture supernatants of rheumatoid synovial fluid mononuclear cells also had a pronounced inhibitory action on IL-l-stimulated thymocyte proliferation (26). In our experiments, the 130- to 150-kDa fraction of supernatants of LPS-stimulated monocytes and EC suppressed not only IL-l-induced thymocyte proliferation, but also the response to IL-2 of CTLL-2 cells and its production by PBMC. The possibility that the inhibition of IL-2 synthesis was due to an IL-l inhibitor of the type described by Lotz et al. is rendered unlikely by the fact that our supernatants did not affect IL-l-stimulated tibroblast proliferation (34, 35). The apparent presence of an IL-2 inhibitor in our preparations and not in the supernatants of Lotz et al. may be due to the additional stimulation provided by LPS in our cultures. Production of an inhibitor of IL-2 synthesis has been described in cultures of adherent cells from lepromatous leprosy patients (27). the macrophage-like cell line U937 (24, 25), and a melanoma cell line (21). In addition, secretion of an inhibitor of the IL-2 response by glioblastoma cells (22) has also been reported. Krakauer has described a MDSF obtained from the human monocytic leukemia cell line, THP-1, which inhibited both the production of IL-2 and its effects on target cells (23). The molecular weight of this factor, eluted from AcA 54 Ultrogel, was 60-70 kDa in contrast to the molecular weight of the 130-150 kDa of our inhibitor eluted from Sephadex G200. Despite this difference in size, the factor described by us cross-reacted with an antibody produced against the MDSF from THP-I, indicating a close relationship between these two agents. Antigen-presenting cells appear to play the additional role of being immunoregulatory cells, as indicated by the capacity of cells of the monocyte-macrophage series to secrete a number of immunosuppressive agents, i.e., PGE, (36.37), IL-l inhibitor (25-27), and the present T cell inhibitor which diminishes IL-2 synthesis and IL-2 action. In view of the fact that EC have also been demonstrated to act as antigen-presenting cells (6, 38, 39) and recent suggestions that antigen presentation may take place on the endothelial surface (40-42), our observation of the secretion of an inhibitor of T cell function by human umbilical vein EC is of special interest. It suggests that modulation of the immune response by an agent secreted by the endothelium may be initiated on the endothelial surface or in the immediate perivascular area.
MACROPHAGE
AND ENDOTHELIAL
CELL-DERIVED
T CELL INHIBITOR
149
ACKNOWLEDGMENTS We thank Dr. Hugo Jasin for his help in the production of IL-2 inhibitor antibody, Dr. Theresa Krakauer for making available a preparation of MDSF from the THP-I cell line, Dr. Druie Cavender for providing us with the cell lines U937 and Bl 1, and Mr. Dale Edelbaum and Mrs. Carla Suson for their excellent assistance in these experiments. This work was supported by NIH Program/Project Grant AMO9989.
REFERENCES 1. Unanue, E. R., The regulatory role of macrophages in antigenic stimulation. Part 2: Symbolic relationship between lymphocytes and macrophages. Adv. Immunol. 1, 136, 1981. 2. Beller, D. I., and Unanue, E. R., Reciprocal regulation of macrophage and T cell function by way of soluble mediators. Lymphokines 6, 25, 1981. 3. Inaba, K., and Steinman, R. M., Accessory cell-T lymphocyte interactions: Antigen-dependent and -independent clustering. 1. Exp. Med. 163, 247, 1986. 4. Nathan, C. F., Secretory products of macrophages. J. C/in. Invest. 79, 319, 1987. 5. Barr, R. S., Goldsmith, J. C., and Owen, W. G.. Properties of endothelial cells in culture: Recent studies. Lymphokines 7, 249, 1982. 6. Hirschberg, H., Bergh, 0. J., and Thorsby, E., Antigen presenting properties of human vascular endothelial cells. J. Exp. Med. 152, 249s 1980. 7. Stem. D. M., Bank, I., Naworth, P. P., Cassimeris, J., Kisie, W., Fenton, W., II, Dinarello, C., Chess, L., and Jaffe, E. A., Self-regulation of procoagulant events on the endothelial cell surface. J. Exp. Med. 162, 1223, 1985. 8. Miossec, P.. Cavender, D., and Ziff, M., Production of interleukin 1 by human endothelial cells, J. Immunol. 136, 2486, 1986. 9. Smith, K. A., Lachman, L. B., Oppenheim, J. J.. and Favata. M. F., The functional relationship of the interleukins. .I. Exp. Med. 151, 1551, 1980. 10. Oppenheim, J. J., and Gery. I., Interleukin 1 is more than interleukin. Immunol. Today 3, 113. 1982. Il. Miossec, P., Kashiwado, T., and Ziff. M., Inhibitor of interleukin 2 in rheumatoid synovial fluid. Arthritis Rheum. 30, 121, 1987. 12. Kashiwado, T.. Miossec, P., Oppenheimer-Marks, N., and Ziff, M., Interleukin 2 inhibitor in rheumatoid synovial fluid. Arthritis Rheum. 30, 1339, 1987. 13. Wolf, R. L., and Andreoni, J., Soluble inhibitory factor (SIF) in normal human serum. Cell. Immunol. 61, 299, 1982. 14. Djeu, J. Y., Kasahara, T., Balow, J. E., and Toskos, G. C., Decreased interleukin 2 inhibitor in sera of patients with autoimmune disorders. Clin. Exp. Immunol. 65, 279, 1986. 15. Nicholas, N. S.. Panayi, G. S., and Touri, A. M., Human pregnancy serum inhibit interleukin 2 production. Clin. Exp. Immunol. 58, 587, 1984. 16. Nath. I., Ellner, J. J., Bouknight, R., Tomford, J. W.. Kleinhenz, M. E., and Edmons. K. L., Interrelationship of immunoregulatory cells and serum factors in sarcoidosis. J. Immunol. 125, 1071, 1980. 17. Siegel, J. P.. Djeu, J. Y., Stocks, N. I., Masur, H., Glemann, E. P., and Quinnan, G. V., Jr., Sera from patients with acquired immunodeficiency syndrome inhibit production interleukin-2 by normal lymphocytes. J. C/in. Invest. 75, 1957, 1985. 18. Dinarello, C. A., Rosenwasser, L. J., and Wolff, S. M., Demonstration of a circulating suppressor factor of thymocyte proliferation during endotoxin fever in human. J. Zmmunol. 127, 2517, 1981. 19. Nicholas, N. S., and Panayi, G. S., Inhibition of IL-2 production by retroplacental sera: A possible mechanism for human fetal allograft survival. Amer. J. Reprod. Immunol. Microbial. 9, 6, 1985. 20. Liao, Z., Grimshaw. R. S., and Rosenstreich, R. S., Identification of a specific interleukin 1 inhibitor in the urine of febrile patients. J. Exp. Med. 159, 126, 1984. 21. Hersey, P., Bindon, C., Czemiecki, M., Spurling, A., Wass, J., and McCarthy, W. H., Inhibition of interleukin 2 production by factors released from tumor cells. J. Immunol. 131, 2837, 1983.
150
KASHIWADO.
OPPENHEIMER-MARKS,
AND
ZIFF
22. Fontana, A., Hengartner, H., Tribolet, N., and Weber, E.. Glioblastoma cells release interleukin 1 and factors inhibiting interleukin 2 mediated effects. J. Immunol. 132, 1837, 1984. 23. Krakauer, T., A macrophage derived factor that inhibits the production and action of interleukin 2. J. Leukocyte Biol. 38, 429, 1985. 24. Fujiwara, H., and Elhter, J. J., Spontaneous production of a suppressor factor by the human macrophage-like cell line U937.1. Suppression of interleukin 1, interleukin 2 and mitogen induced blastogenesis in mouse thymocytes. J. Immunol. 136, 181, 1986. 25. Fujiwara, H., Toossi, Z., Ohnishi, K., Edmonds, K., and Ellner, J. J., Spontaneous production of a suppressor factor by a human macrophage-like cell line U937. II. Suppression of antigen- and mitogen-induced blastogenesis. IL-2 production and IL-2 receptor expression in T lymphocytes. J. Immunol.
138,
197,
1987.
26. Lotz, M., Tsoukas, C. D., Robinson, C. A., Dinarello, C. A., Carson, D. A., and Vaughan, J. H., Basis for defective responses of rheumatoid arthritis synovial fluid lymphocytes to anti-CD3(T3) antibodies. J. Clin. Invest. 78, 713, 1986. 27. Nath, I., Jayaraman, J., Sathis, M., Bhutani, L. K., and Sharama, A. K., Inhibition of interleukin 2 production by adherent cell factors from lepromatous leprosy patients. Clin. Exp. Zmmunol. 58, 531, 1984. 28. Sandborg, C. I., Berman, M. A.. Andrews, B. S., Mirick, G. R., and Friou, G. J.. Increased production of an interleukin 1 (IL-l) inhibitor with Bbroblast stimulating activity by mononuclear cells from patients with scleroderma. Clin. Exp. Immunol. 66, 312, 1986. 29. Jaffe, E. A., Nachman, R. L., Becker, C. G., and Minick, C. R., Culture of human endothelial cells derived from umbilical veins. J. Clin. Invest. 52, 2745, 1973. 30. Smith, K. A., Gilbride, K. J., and Favata, M. F., Lymphocyte activating factor promotes T-cell growth factor production by cloned murine lymphoma cells. Nature (London) 287, 853, 1980. 3 1. Oppenheim, J. J., Stadler, B. M., Siraganian, R. P., Mage, M., and Mathieson, B., Lymphokines: Their role in lymphocyte responses. Properties of interleukin 1. Fed. Proc. 41, 257, 1982. 32. Caplan, B , and Rothenberg, E., High-level secretion of interleukin 2 by a subset of proliferating thymic lymphoblasts. J. Immunol. 133, 1983, 1984. 33. Lotz, M., Tsoukas, C. D., Fong, S., Dinarello, C. A., Carson, D. A., and Vaughan, J. H., Release of lymphokines after infection with Epstein-Barr virus in vitro. II. A monocyte-dependent inhibitor of interleukin 1 downregulates the production of interleukin 2 and interferon-y in rheumatoid arthritis. J. Immunol. 136, 3643, 1986. 34. Postlethwaite, A. E., Lachman, L. B., and Kang, A. H., Induction of fibroblast proliferation by interleukin 1 derived from human monocytic leukemia cells. Arthritis Rheum. 27, 995, 1984. 35. Kom, J. H.. Torres, D., and Downie. D., Clonal heterogeneity in the fibroblast response to mononuclear cell derived mediators. Arthritis Rheum. 27, 174, 1984. 36. Goodwin, J. S., Bankhurst, A. D., and Morley, J., Suppression of human T-cell mitogenesis by prostaglandin. J. Exp. Med. 146, 1719, 1977. 37. Farrar, W. L., and Humes, J. L.. The role of arachiodonic acid metabolism in the activities of interleukin 1 and 2. J. Immunol. 135, 1153, 1985. 38. Ashida, E., Johnson, A., and Lipsky, P. E., Human endothelial cell-lymphocyte interactions. Endothelial cells function as accessory cells necessary for mitogen-induced T lymphocyte activation in vitro. J. Clin. Invest. 671, 1490, 1981. 39. Nunez, G., Ball, E. J., and Stastny, P., Accessory cell function of human endothehal cells, I. A subpopulation of Ia positive cells is required for antigen presentation. J. Zmmunol. 131, 666, 1983. 40. Pober, J. S., Gimbrone, M. A., Jr., Collins, T., Cotran, R. S., Ault, K. A., Fiers, W., Krensky, A. M., Clayberger, C., Reiss, C. S., and Burakoff, S. J., Interactions of T lymphocytes with human vascular endothelial cell surface antigens. Immunobiology 168, 483, 1984. 41. Wagner, C. R., Vetto, R. M., and Burger, D. R., Expression of I-region-associated antigen (Ia) and interleukin 1 by subcultured human endothelial cells. Cell. Zmmunol. 93, 91, 1985. 42. Geppert, T. D., and Lipsky, P. E., Antigen presentation by interferon-y-treated endothelial cells and fibroblasts: Differential ability to function as antigen-presenting cells despite comparable Ia expression. J. Immunol. 135, 3750. 1985. Received February 27, 1989; accepted May 16, 1989