Development of a simple, sensitive and specific bioassay for interleukin-1 based on the proliferation of RPMI 1788 cells comparison with other bioassays for IL-1

Journal of Immunological Methods, 135 (1990) 25-32

25

Elsevier JIM05762

Development of a simple, sensitive and specific bioassay for interleukin-1 based on the proliferation of RPMI 1788 cells Comparison with other bioassays for IL-1 Peter Vandenabeele, Wim Declercq, Claude Libert and Walter Fiers Laboratory of Molecular Biology, State University of Ghent, Ghent, Belgium

(Received 2 April 1990, revised received 6 July 1990, accepted 28 August 1990)

The IL-l-dependent proliferation of R P M I 1788, a human EBV-transformed cell fine, was used to develop a biological assay system for IL-1. Preparations of r h I L - l a and rhIL-1B, as well as r m I L - l f l exhibited a specific biological activity (50% of the maximal response) between 5.8 x 10 s and 8.6 x 108 U / m g . Remarkably, a 3-5-fold reduced specific biological activity was noticed for r m - I L - l a , viz. 1.7 x l0 s U / m g . The IL-l-dependent proliferation of R P M I 1788 cells was compared with other IL-1 test systems, such as the IL-l-mediated induction of IL-2 in EL4-NOB-1, LBRM-33-1A5 and thymocytes, and the IL-l-driven induction of cytotoxic activity by PC60 cells, the so-called CIA assay. The cytokine-dependent growth of R P M I 1788 cells is highly specific for IL-1, and no other cytokine tested induced a proliferative response. The presence of high concentrations of r m T N F , r h T N F or rhIL-6 did not interfere with the quantification of IL-1. Additionally, we evaluated the detection of IL-1 in the presence of mitogens, phorbol ester or calcium ionophore, as well as the determination of IL-1 in serum and PF samples of human and murine origin. Key words: Interleukin-1; Bioassay; Cytokine

Correspondence to: W. Fiers, Laboratory of Molecular Bi-

ology, State University of Ghent, Ledeganckstraat 35, B-9000 Ghent, Belgium. Abbreviations: A23186, calcium ionophore; ConA, concanavalin A; CSF, colony-stimulatingfactor; CPE, cytopathic effect; EBV, Epstein-Barr virus; EGF, epidermal growth factor; G-CSF, granulocyte CSF; GM-CSF, granulocyte-macrophage CSF; h, human; IL, interleukin; LAF, leukocyte-activating factor; LPS, lipopolysaccharide; m, murine; M-CSF, macrophage CSF; PDGF, platelet-derivedgrowth factor; PF, peritoneal fluid; PHA, phytohemagglutinin; PWM, pokeweed mitogen; r, recombinant; SI, stimulation index; SD, standard deviation; TGF-fl, transforming growth factor-r; TNF, tumor necrosis factor; TPA, 12-O-tetradecanoylphorbol 13-acetate.

Introduction A specific and sensitive detection system for biologically active IL-1 is particularly useful for the determination of possible IL-1 induction under pathological conditions or after in vivo administration of cytokines during therapy. Several such IL-1 bioassays have been developed. The classical thymocyte co-stimulation or L A F assay (Gery et al., 1972), the LBRM-33-1A5 assay (Conlon, 1983) and the EL-4 NOB-1 assay (Gearing et al., 1987) are based on the IL-l-mediated induction of IL-2 production. The CIA assay (Erard et al., 1984) makes use of the IL-l-dependent differentiation of

0022-1759/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

26 a rat-mouse T cell hybridoma PC60 to cytotoxic T cells. IL-1 has also been measured by its cytostatic activity on the human melanoma A375 (Nakano et al., 1987), its mitogenic effect on a subclone of the murine D10 T cell line (Hopkins and Humphreys, 1989) and on a fibroblast cell line (Schmidt et al., 1982), and by its capacity to induce IL-6 secretion in the human MG-63 osteosarcoma cell line (Van Damme and Van Snick, 1987). Previously (Vandenabeele et al., 1988), we demonstrated that IL-la is the autocrine growth factor of RPMI 1788, an EBV-transformed cell line. In this report, we describe conditions for a sensitive and specific IL-1 detection system based on the proliferation of low density-seeded RPMI 1788 cells. Furthermore, we compared the LAF, NOB-l, LBRM, CIA and RPMI assays in their sensitivity to murine and human IL-la and IL-1/3 preparations and in their response to other cytokines.

Materials and methods

Cytokines Preparations of rhlL-la, rhlL-1/3, rmlL-la and rmlL-1/3, purified to homogeneity, were a gift from Dr. A. Shaw (Glaxo IMB, Geneva, Switzerland), rmlL-la and rmlL-1/3 were also provided by Dr. P. Lomedico (Hoffmann-La Roche, Nutley, NJ, U.S.A.). Protein content was estimated by the Amido-Schwarz method (Schaffner and Weissmann, 1973) with bovine serum albumin as a standard, rhlL-2 (4.4 × 106 U/mg, CTLL-2 assay (Gillis et al., 1978)) was provided by Biogen S.A. (formerly at Geneva, Switzerland). rmlL-4 (2000 U/ml, CTLL-2 assay (Miiller and Vandenabeele, 1987)) and rhlL-4 (1000 U/ml, proliferation of TPA stimulated, small, resting B cells from peripheral blood, personal communication by Dr. J. Gordon, University of Birmingham, U.K. (Jayaram et al., 1989)) were provided by Dr. R. Devos (Roche Research Ghent, Ghent, Belgium). rmlL-5 (2.0 x 107 U/mg, BCL1 assay (Swain et al., 1982)), rhlL-5 (5 × 10 6 U/mg, eosinophil peroxidase assay (Strath et al., 1985)), rmTNF (3.3 x 109 U/mg, WEHI-164 assay (Espevik and Nissen-Meyer, 1987)) and rhTNF (1.7 × 108 U/mg, WEHI-164 assay) were supplied by Dr. J. Tavernier (Roche

Research Ghent). rmIL-6 (8 × 108 U/mg, 7TD1 assay (Van Snick et al., 1986)) was kindly provided by Dr. J. Van Snick (Ludwig Institute for Cancer Research, Brussels, Belgium). rhIL-6 (2 × 108 U/mg, 7TD1 assay) and rmIFN-y (2 × 107 U/mg, CPE-inhibition assay on L929 cells (Derynck et al., 1980)) were prepared in our laboratory by Y. Guisez. rhGM-CSF, rmGM-CSF (2.5 × 108 U/mg, FDCpl assay (DeLamarter et al., 1985)) and rhG-CSF (4 × 107 U/rag, DA2 assay (Moreau et al., 1987)) were obtained from Dr. J. DeLamarter (Glaxo IMB). rhPDGF AA and rhPDGF BB were kindly provided by Dr. C.-H. Heldin (Ludwig Institute for Cancer Research, Uppsala, Sweden). Murine EGF was obtained from Boehringer (Mannheim, F.R.G.) and hTGF/3 was purchased from Paesel (Frankfurt, F.R.G.). Conditioned medium from WEHI-3 cells was used as a source of mIL-3 (2300 U / m l , Ea3.15 assay (Pallacios et al., 1984)) and supernatant from L929 cells as a source of mM-CSF (not quantified). In the various cytokine tests, 1 U was defined as the amount of factor capable of inducing 50% of the maximal signal.

Cell lines and cell culture RPMI 1788, an EBV-transformed human B cell line, LBRM-33-1A5, a murine T cell lymphoma, and S194, a murine myeloma cell line, were obtained from ATCC (Rockville, MD, U.S.A.). EL-4 NOB1, a subclone of the murine thymoma EL-4, was provided by Dr. A.J.H. Gearing (NIBSC, London, U.K.). PC60 is a hybrid between a rat thymoma and a murine cytolytic T cell line (Erard et al., 1984) and was kindly donated by Dr. M. Nabholz (ISREC, Epalinges, Switzerland). RPMI 1788 cells were split 1 : 10 twice a week, without the addition of an IL-1 source, since the endogenously produced IL-1 at higher cell densities (> 103/ml) was sufficient for adequate cell growth. All cell cultures and assays were performed in RPMI 1640 culture medium, supplemented with 10% fetal calf serum, 50 U / m l penicillin 'G', 50 # g / m l streptomycin sulfate, 2 mM L-glutamine, 5 X 10 -5 M fl-mercaptoethanol, 1 mM sodium pyruvate and 10 mM Hepes. PC60 was cultured in DMEM with the same additives, supplemented with 10/zg/ml folic acid, 20/~g/ml L-arginine and 36 # g / m l t-asparagine.

27 IL-1 assays Thymocytes for the LAF-assay (Gery et al., 1972) were isolated from 4 to 10 weeks old C3HHeJ mice (Iffa-Credo, Saint Germain-sur-l'Arbresle, France) and were cultured at 2 × 1 0 6 cells/well in the presence of 2.5 # g / m l PHA (Wellcome, Dartford, U.K.). IL-l-mediated proliferation was measured by [3H]thymidine incorporation during the last 18 h of a 3 day incubation. The LBRM-33-1A5 conversion assay (Conlon, 1983) was performed as follows: cells treated with mitomycin C (Sigma Chemical Co., St. Louis, MO, U.S.A.) were incubated at 5 × 104 cells/well containing serially diluted IL-1 samples in the presence of 2 . 5 / t g / m l PHA. The next day, CTLL2 cells (2 × 1 0 4 cells/well) were added to quantify the IL-l-induced IL-2 secretion. CTLL-2 proliferation was determined by [3H]thymidine incorporation during the last 4 h of a 24 h incubation period. Essentially the same protocol was followed for the EL-4 NOB-1 assay (Gearing et al., 1987), subsequently referred to as the NOB-1 assay; i.e., incubation of 2 × 1 0 4 t k - EL-4 NOB-1 cells/well with serial dilutions of IL-1 and measurement of the secreted IL-2 by addition of 4 × 103 CTLL-2 cells/well. IL-1 was also measured by its ability to induce cytotoxicity in PC60 cells in the presence of IL-2,

I N F L U E N C E O F CELL DENSITY O N rh I L - I # - D E P E N D E N T

the so-called CIA assay (Erard et al., 1984). Briefly, 5 x 103 PC60 cells/well were cultured in medium containing 60 U / m l IL-2 and serial dilutions of test material in 96 well round bottom microtiter plates. After 3 days, 5 x 103 5~Cr-labelled $194 myeloma target cells and 10 /~g/ml ConA were added. After 5 h specific 51Cr release was determined. The day before use in the assay, a saturated culture of RPMI 1788 cells (approximately 1 0 6 cells/ml) was split 1 : 3 in order to obtain high cell viability. Thoroughly washed cells were then incubated in flat-bottomed 96 well microtiter plates in a final volume of 200 t~l/well at 2000, 1000, 500 or 200 cells/well over 3, 4, 5, or 6 days, respectively. Proliferation was measured by the addition of [3H]thymidine (0.5 /~Ci/well) during the last 5 h of the incubation. Before adding the [3H]thymidine, we evaluated the IL-l-dependent response by microscopic examination. A good response was characterized by the formation of large clumps of viable RPMI 1788 cells. If the IL-1mediated clustering was still weak, incubation was continued. When not mentioned explicitly, we used standard assay conditions, viz., 500 cells/well over 5 days and a 5 h pulse with [3H]thymidine. Addition of 1 m M sodium pyruvate to the culture medium was absolutely required for clear-cut ILl-dependent proliferation (data not shown).

PROLIFERATION

'o x E

15 50

•

6/o/°~°\o /-

40

15

]

10

10

~

o

50 o c

i/~.i-.i-i

/o

20

1__0/07 E I i

10

O's---O/

~ /

i .jl/"

"6 E

i/i

o

A--A--A~A 0.01

0.1

1

10

100

Concentretion of rh I L - l f l (pg/rnl)

200

500

1000

2000

Number of cells per well

Fig. 1. Effect of RPMI 1788 cell density on IL-1 detection. The RPMI assay was performed at 200 (A), 500 (z~),1000 (O) or 2000 cells/well (o) during 6, 5, 4 or 3 days, respectively.A shows the IL-l-dependent proliferation at different cell concentrations.In B, the specific biological activities (o) and stimulation indices (SI) (O) were calculated from the data shown in A. Both values are expressed as a function of the number of ceils/well. Bars represent the SD for three measurements.

28

Results

TABLE

SPECIFIC

BIOLOGICAL

Influence of cell density on the detection of IL-1 Since the growth of RPMI 1788 cells depends on autocrine-acting IL-lc~ (Vandenabeele et al., 1988), we studied the possibility of using the growth of these cells as an IL-1 assay. Initially, we investigated the influence of cell density on the proliferation of RPMI 1788 in the presence of exogenously added IL-1 (Fig. 1A). In this experiment, the proliferative response was evaluated by means of the stimulation index (SI), i.e., the ratio between maximal signal and blank level, and by the specific biological activity (U/mg). The former value is a good measure of the reliability and the latter of the sensitivity of a bioassay. These values were calculated from data shown in Fig. 1A and are presented in Fig. lB. The specific biological activity is directly proportional to the number of cells per well. Conversely, the SI is reciprocally related to cell density. Likewise, the SD values on the calculation of the specific biological activities increase with lower numbers of cells/well. In the following experiments, the RPMI assay was performed over 5 days at 500 cells/well.

MURINE

IL-la

TABLE

I

AND

ACTIVITIES IL-lfl

IN

OF HUMAN

VARIOUS

ASSAY

OF THE

RPMI

ASSAY

AS COMPARED

The mean values of three determinations are given (SD between brackets). The specific biological activity was calculated by plotting the data as shown for the RPMI assay in F i g . 1 A . Identical IL-1 preparations were used in the different I L - 1 assays. Specific biological activities ( × 1 0 - 7 U / r a g ) RPMI

CIA

(6)

5.9

(0.4)

NOB-1

LAF

LBRM

rmlL-la

17

75(11)

2.8 (0.7)

460 (32)

rhlL-la

58 ( 1 4 )

0.13 ( 0 . 0 2 )

65 ( 1 1 )

2.4 (0.6)

580 (97)

rmlL-lfl

86 ( 3 4 )

6.3

(0.4)

71

(9)

2.0 (0.5)

2 7 0 (38)

rhlL-lfl

65 ( 3 2 )

4.7

(0.6)

98

(8)

1.7 (0.6)

360 (68)

Sensitivity and specificity of the RPMI assay and comparison with other IL-1 test systems To investigate the sensitivity of the RPMI assay, we used preparations of rhlL-la, rhlL-lfl, rmlL-la and rmlL-lfl purified to homogeneity. To check their biological activities, several other IL-l-dependent bioassays were performed. Table I presents the specific biological activities of identi-

WITH

OTHER

IL-1 ASSAYS

RPMI

CIA

NOB-1

LAF

LBRM

rhlL-2

100 *

-

-

+

+

+

mlL-3

2,300 * *

.

rmlL-4

2,000 * *

-

+

+

rhlL-4

1,000 * *

.

.

.

.

.

.

.

. .

-

. +

rmlL-5

10 *

.

.

.

.

rhlL-5

5,000 *

.

.

.

.

rmlL-6

10 *

-

-

-

+

-

rhlL-6 rmTNF

10 * 500 *

-

7.9 x 104

2.0 x 106

+ +

4.2 X 103

rhTNF

5,000 *

.

10 * 100 *

. .

rmGM-CSF rhGM-CSF

.

. . .

.

. . .

ND . .

. .

rhG-CSF

100 *

.

.

.

.

.

mM-CSF rmlFN-y

ND 100 *

. -

.

.

.

.

100 *

.

.

.

.

.

mPDGF

AA

mPDGF

BB

-

6.6 X 102

-

100 *

.

.

.

.

.

mEGF

100 *

.

.

.

.

.

hTGF-fl

100 *

.

.

.

.

.

Key."

-,

SYS-

TEMS

II

SPECIFICITY

AND

undetectable activity; + , activity; *, concentration in n g / m l ;

* *,

concentration in U / m l ;

-

ND,

not determined.

29

cal samples of murine and human IL-lct and IL-lfl in various IL-1 assays. The IL-1 assays examined reached 50% of the maximal signal at the following IL-1 concentrations: approximately 250 fg/ml (LBRM), 1.5 pg/ml (RPMI and NOB1), 20 pg/ml (CIA) and 100 pg/ml (LAF). In the NOB-l, LBRM or LAF assays no significant distinction was observed between the specific biological activities of IL-lct and IL-lfl from either species. Remarkably, IL-la exhibited some species specificity in the human RPMI and in the rat/mouse CIA assay. In several experiments and using different preparations of rmIL-la (data not shown), we found that rmIL-la displayed a 3-5fold lower specific biological activity than rhIL-lct on the human RPMI 1788 cell line. On the contrary, the same rmIL-la preparation showed a 50-fold higher specific biological activity than rhIL-la on PC60 cells. The specificity of various IL-1 assays was examined using a series of human and murine cytokine preparations (Table II). The CIA, NOB-l, LAF and LBRM assays respond to mTNF, although in most cases a high concentration of the latter cytokine is required. Furthermore, since CTLL-2 is used for the detection of IL-l-induced IL-2 secretion in LBRM and NOB-1 cells, murine and human IL-2 and murine IL-4 also generate a positive signal (Miiller and Vandenabeele, 1985), unless preincubation steps are introduced (Gearing et al., 1987). The thymocyte co-stimulation assay, which is sensitive to mTNF, IL-2 and mIL-4, also responds to IL-6 (Le et al., 1988). Since the induction of IL-1 secretion is often accompanied by enhanced TNF and/or IL-6 production and since many pathological conditions are characterized by elevated levels of TNF (Fiers et al., 1989) a n d / o r IL-6 (Le and Vilcek, 1989), we examined the effect of high doses of TNF and IL-6 on the quantification of IL-1 in the RPMI assay. As shown in Fig. 2, the presence of rhTNF a n d / o r rhlL-6 did not affect the IL-l-dependent proliferation of RPMI 1788 cells. At high concentrations IL-6 generated some minor growth-enhancing activity. This IL-6-mediated proliferation completely dropped to background level at IL-6 concentrations below 1 ng/ml (data not shown). Therefore we may conclude that the RPMI assay is strictly IL-1 specific.

INFLUENCE OF TNF"AND IL-6 ON IL-1 -DEPENDENT PROUFERATION I

o x

E 30 P o i.

20

o

10 E I

0.1

~

~'0

io0

l:llt

*

Fig. 2. Effect of T N F and IL-6 on IL-1 quantification. RPMI cells were cultured at 500 ceils/well during 5 days. rhIL-lfl (©), rhIL-lfl +500 n g / m l r m T N F (e), rhIL-lfl +200 n g / m l r h T N F (4), rhIL-lfl + 100 n g / m l rhIL-6 (1), rhIL-lfl + rhTNF + rhIL-6 (rT) were serially diluted. Bars represent proliferation levels in the absence of IL-1 when T N F a n d / o r IL-6 were given alone.

Influence of mitogens, phorbol ester and calcium ionophore on the quantification of IL-1 Many cytokine preparations originate from stimulated cell cultures. Hence, we studied the influence of ConA, PHA, PWM, TPA, calcium ionophore, TPA + calcium ionophore, or LPS on the proliferation of RPMI 1788 cells in the presence of saturating amounts of rhIL-lfl (Table III, second column). CortA, PWM and calcium ionophore hardly affected the level of rhIL-lfl-mediated growth. TPA, TPA + calcium ionophore and LPS slightly decreased the plateau level proliferation. PHA, on the other hand, was strongly inhibitory. At a dilution of 1/50, these agents no longer influenced the IL-l-mediated proliferation (data not shown). As indicated in Table III (third column), none of the mitogens altered the biological quantification of 100 pg/ml rhIL-lfl. Determination of rhlL-lfl in serum and PF Table IV shows data on the measurement of 100 pg/ml rhIL-lfl in serum and in PF of human and routine origin. Although the presence of body fluids at a 1/10 dilution partially reduced the maximal IL-l-dependent proliferation, the biological quantification of 100 pg/ml rhIL-lfl was not influenced in most cases. We were not able to

30 TABLE III INFLUENCE OF MITOGENS, PHORBOL ESTER AND CALCIUM 1ONOPHORE ON THE QUANTIFICATION OF rhlL-lfl Influence of ConA (5 /zg/ml, Pharmacia), TPA (20 ng/ml, Sigma), PHA (10 /~g/ml, Wellcome), PWM (1/500, Gibco), A23186 (50 ng/ml, Sigma), or LPS (10/~g/ml, Sigma) on the proliferation of RPMI 1788 cells after 1/10 dilution in the absence (first column, cpm) or presence of rhlL-lfl (10 pg/ml, second column, cpm). In the third column is shown the biological activity (U/ml) of serial dilutions of 100 pg/ml rhlL-1/3 spiked with the mitogens at the concentrations mentioned above. Control wells contained only rhlL-lfl in culture medium. Values are given as the mean of three measurements (SD between brackets).

A375 assay (Nakano et al., 1987), also responded to m T N F (Table II). These assays revealed at least a 100-fold difference in specific biological activity between IL-1 and mTNF. However, the response to the latter might interfere with the quantification of IL-1; e.g., the measurement of serum IL-1 levels after LPS administration or in the course of therapeutic m T N F treatment in mice (Libert et al., in preparation). The IL-1 quantification in the RPMI assay was not disturbed by high doses of TNF a n d / o r IL-6. Recently, IL-6 was reported to

TABLE IV Mitogen

Control ConA TPA PHA PWM A23186 TPA + A23186 LPS

1/10 diluted no IL-1

1/10 diluted 10 pg/ml rhIL-lfl

U / m l (100 pg/ml rhIL-lfl)

3,805 (1,081) 2,926 (951) 4,221 (745) 2,156 (84) 3,817 (639) 3,370 (362)

34,057(2,911) 36,504(3,712) 20,108(1,372) 6,845 (427) 31,381(7,005) 30,564(2,643)

90 (17) 67 (7) 107 (24) 121 (39) 54 (8) 86 (12)

7,667(1,479) 3,630 (1,143)

19,530(1,641) 16,898(5,013)

83 (14) 57 (19)

prevent this reduction by chloroform extraction (Canon et al., 1989) (data not shown).

Discussion In this paper, we show that the IL-l-dependent proliferation of RPMI 1788 cells can be used as a sensitive and highly specific bioassay for IL-1. The RPMI assay, achieving half-maximal proliferation at about 1.5 pg/ml, except for mlL-la, was as sensitive as the NOB-1 and the D10 assays (Helle et al., 1988), but approximately ten times less sensitive than the LBRM conversion assay. The CIA and LAF assays are 10 and 100 times less sensitive, respectively, than the RPMI test. Except for some minor activity by rhlL-6, no other cytokine besides IL-1 was able to promote the growth of RPMI 1788. This strict specificity is only equalled by the D10 assay if performed under conditions as described by Hopkins and Humphreys (1989). Most IL-1 assays, including the

DETECTION OF rhlL-lfl IN SERUM OR IN PF Human sera were derived from healthy volunteers. PF samples were obtained from continuous ambulatory peritoneal dialysis patients with no sign of peritoneal inflammation. Murine serum samples were obtained by cardiac puncture of C57BL/6 mice, PF by i.p. injection of 3 ml PBS, followed after 2 h by collection of the fluid. Serum and PF samples were assayed in the absence (first column) and presence of 10 pg/ml rhlL-lfl (second column). In the third column we show the biological activity (U/ml) of serial dilutions of 100 pg/ml rhlL-lfl added to serum and PF samples. All samples were tested in threefold (SD between brackets). 1/10 diluted no IL-1

1/10 diluted 10 pg/ml rhlL-lfl

4,236

(561)

32,806 (2,625)

64 (8)

Human samples Serum 1 4,518 (308) Serum 2 4,072 (827) Serum 3 4,556 (1,052) Serum 4 3,768(1,013) Serum 5 5,335(1,943) PF 1 5,626 (466) PF 2 3,347 (213) PF 3 3,441 (555) PF 4 2,167 (343) PF 5 3,473 (874)

37,088 (1,870) 27,269 (10,362) 34,872 (3,138) 30,938 (2,784) 27,392 (1,370) 30,508 (3,966) 32,805 (5,249) 24,605 (5,682) 20,471 (3,480) 28,069 (2,526)

63 (23) 69 (21) 81 (15) 69 (17) 120 (17) 51 (17) 62 (15) 63 (14) 30 (9) 67 (9)

Murine samples Serum 1 4,834 (708) Serum 2 6,989 (930) Serum 3 7,945(1,641) Serum 4 4,222 (1,084) Serum 5 4,330 (806) PF 1 5,111 (1,102) PF 2 3,460 (254) PF 3 4,414 (684) PF 4 6,537 (2,514) PF 5 3,577 (535)

24,225 23,005 24,533 25,100 22,920 21,127 32,346 27,580 26,557 31,033

Control

(1,453) (920) (981) (2,008) (2,980) (5,282) (2,264) (4,414) (6,905) (3,414)

U / m l (100 pg/ml rhlL-lfl)

50 (10) 91 (27) 59 (20) 68 (19) 65 (13) 51 (9) 85 (20) 88 (17) 60 (13) 88 (20)

31 sustain the growth of low density-seeded R P M I 1788 ceils (Yokoi et al., 1990). However, the data presented in that paper do not exceed the very low proliferative activity of rhlL-6 also observed by us (Fig. 2). The R P M I assay, like the LAF assay, exhibited substantial variability, as illustrated by the high SD values when the specific biological activities were calculated (Table I). Therefore, it is advisable to perform triplicate measurements for each test sample. In several experiments, a five-fold difference in specific biological activity between r m l L - l a and r h I L - l a on the proliferation of R P M I 1788 cells was observed. This might reflect an element of species specificity for I L - l a on human B cells. The fact that the same human and murine I L - l a preparations tested on LBRM, NOB-1 or thymocytes did not display this distinction, excludes the possibility of an artefact. This differential response between R P M I 1788, on the one hand, and LBRM, NOB-1 and thymocytes, on the other hand, might be due to the presence of different IL-1 receptors on B cells as compared to T cells (Bomsztyk et al., 1989; Chizzonite et al., 1989). Conflicting data are found in the literature concerning the 50-fold discrepancy between the activity of r h l L - l a and other types of IL-1 on the r a t / m o u s e hybridoma PC60. The rhIL-lo~-mediated induction of GM-CSF, IL-6 and IFN-T secretion by PC60 cells also revealed this reduced specific activity (Vandenabeele et al., 1990). However, the IL-l-mediated induction of human IL-2 receptor e¢ (Tac antigen) in transfected PC60 cells containing the human I L - 2 R a gene under control of a SV40 promoter, did not show any difference in specific biological activity between r h I L - l a and rhIL-1]~ (Plaetinck et al., 1989). This suggests that the reduced specific biological activity of r h l L - l a as compared with the other types of IL-1 is not exerted at the level of binding to the IL-1 receptor, but might be located downstream. The same mechanism might account for the distinction in specific biological activity between r m I L - l a and r h I L - l a on human R P M I 1788 cells. The R P M I assay also allowed us to quantify 100 p g / m l rhIL-1]~ in serum and PF as well as in samples containing mitogens, phorbol ester or calcium ionophore. Under some of these condi-

tions lower concentrations of IL-1 will be underestimated due to inhibitory effects. Nevertheless, we believe that the specificity and high sensitivity of the R P M I assay, like the D10 assay, is very useful for unambiguous IL-1 quantification, especially in samples containing several other cytokines.

Acknowledgements We thank Drs. J. DeLamarter, R. Devos, A.J.H. Gearing, M. Goldman, J. Gordon, Y. Guisez, C.-H. Heldin, P. Lomedico, M. Nabholz, A. Shaw, J. Tavernier, J. Van Snick and P. Wingfield for providing purified recombinant cytokines, cell lines, detailed protocols for cytokine assays and human PF samples. We are grateful to W. Burm for technical, and to M. Vandecasteele and W. Drijvers for editorial assistance. P. Vandenabeele is a research fellow with the F G W O , W. Declercq with the OAA and C. Libert with the N F W O . Research was supported by grants from the Belgian Ministry of Science Policy (OAA and I U A P ) and from the FGWO.

References Bomsztyk, K., Sims, J.E., Stanton, T.H., Slack, J., McMahan, C.J., Valentine, M.A. and Dower, K.S. (1989) Evidence for different interleukin 1 receptors in murine B- and T-cell fines. Proc. Natl. Acad. Sci. U.S.A. 89, 8034. Cannon, J.G., Van der Meer, J.W.M., Kwiatkowski, D., Endres, S., Lonnemann, G., Burke, J. and Dinarello, C. (1988) Interleukin-lfl in human plasma: optimization of blood collection, plasma extraction and radioimmunoassaymethods. Lymphokine Res. 7, 457. Chizzonite, R., Truitt, T., Kilian, P.L., Stern, A.S., Nunes, P., Parker, K.P., Kaffka, K.L., Chua, A.O., Lugg, D.K. and Gubler, U. (1989) Two high-affinity interleukin 1 receptors represent separate gene products. Proc. Natl. Acad. Sci. U.S.A. 86, 8029. Conlon, P.J. (1983) A rapid biologic assay for the detection of interleukin 1. J. Immunol. 131, 1280. DeLamarter, J.F., Mermod, J.-J., Liang, C.-M., Eliason, J.F. and Thatcher, D.R. (1985) Recombinant murine GM-CSF from E. coli has biological activity and is neutralized by a specific antiserum. EMBO J. 4, 2575. Derynck, R., Remaut, E., Saman, E., Stanssens, P., De Clercq, E., Content, J. and Fiers, W. (1980) Expression of human fibroblast interferon gene in Escherichia coli. Nature 287, 193.

32 Erard, F., Corthrsy, P., Smith, K.A., Fiers, W., Conzelmann, A. and Nabholz, M. (1984) Characterization of soluble factors which induce the cytolytic activity and the expression of T-cell growth factor receptors of a T-cell hybrid. J. Exp. Med. 160, 584. Espevik, T. and Nissen-Meyer, J. (1986) A highly sensitive cell line, WEHI 164 clone 13, for measuring cytotoxic factor/tumor necrosis factor from human monocytes. J. Immunol. Methods 95, 99. Fiers, W., Beyaert, R., Brouckaert, P., Everaerdt, B., Libert, C., Suffys, P., Takahashi, N., Vanhaesebroeck, B. and Van Roy, F. (1990) Mechanism of action of tumor necrosis factor and its implications for synergizing and antagonizing drugs. Med. Sci., in press. Gearing, A.J.H., Bird, C.R., Bristow, A., Poole, S. and Thorpe, R. (1987) A simple sensitive bioassay for interleukin-1 which is unresponsive to 103 U / m l of interleukin-2. J. Immunol. Methods 99, 7. Gery, I., Gershon, R.K. and Waksman, B.H. (1972) Potentiation of the T-lymphocyte response to mitogens. I. The responding cell. J. Exp. Med. 136, 128. Gillis, S., Ferm, M.M., Ou, W. and Smith, K.A. (1978) T-cell growth factor: parameters of production and a quantitative microassay for activity. J. Immunol. 120, 2027. Helle, M., Boeije, L. and Aarden, L.A. (1988) Functional discrimination between interleukin 6 and interleukin 1. Eur. J. Immunol. 18, 1535. Hopkins, S.J. and Humphreys, M. (1989) Simple, sensitive and specific bioassay of interleukin-1. J. Immunol. Methods 120, 271. Jayaram, B., Devos, R., Guisez, Y. and Fiers, W. (1989) Purification of human interleukin-4 produced in Escherichia coli. Gene 79, 345. Le, J. and Vilcek, J. (1989) Interleukin 6: a multifunctional cytokine regulating immune reactions and the acute phase protein response. Lab. Invest. 67, 588. Le, J., Fredrickson, G., Reis, L.F.L., Diamantstein, T., Hirano, T., Kishimoto, T. and Vilcek, J. (1988) Interleukin 2-dependent and interleukin 2-independent pathways of regulation of thymocyte function by interleukin 6. Proc. Natl. Acad. Sci. U.S.A. 85, 8643. Moreau, J.-F., Bonneville, M., Godard, A., Gascan, H., Gruart, V., Moore, M.A. and Soulillou, J.P. (1987) Characterization of a factor produced by human T-cell clones exhibiting eosinophil-activating and burst-promoting activities. J. Immunol. 138, 3844. Miiller, W. and Vandenabeele, P. (1987) A T-cell clone which responds to interleukin 2 but not to interleukin 4. Eur. J. Immunol. 17, 579. Nakano, K., Okugawa, K., Hayashi, H., Abe, S., Sohmura, Y.

and Tsuboi, T. (1987) Establishment of Dye-Uptake Method (A375 Assay) for quantitative meastirement of IL-I: correlation with LAF assay. Dev. Biol. Stand. 69, 93. Pallacios, R., Henson, G., Steinmetz, M. and McKearn, J.P. (1984) Interleukin-3 supports growth of mouse pre-B-cell clones in vitro. Nature 309, 126. Plaetinck, G., Combe, M.C., Corthrsy, P., Seckinger, P. and Nabholz, M. (1989) Intefleukin 1 and tumor necrosis factor enhance transcription from the SV40 early promoter in a T-cell line. Eur. J. Immunol. 19, 897. Schaffner, W. and Weissmann, C. (1973) A rapid, sensitive and specific method for the determination of protein in dilute solution. Anal. Biochem. 56, 502. Schmidt, J.A., Mizel, S.B., Cohen, D. and Green, I. (1982) Interleukin-1, a potent regulator of fibroblast proliferation. J. Immunol. 128, 2177. Strath, M., Warren, D.J. and Sanderson, C.J. (1985) Detection of eosinophils using an eosinophil peroxidase assay. Its use as an assay for eosinophil differentiation factors. J. Immunol. Methods 83, 209. Swain, S.L. and Dutton, R.W. (1982) Production of a B-cell growth-promoting activity, (DL)BCGF, from a cloned Tcell line and its assay on the BCL 1 B-cell tumor. J. Exp. Med. 156, 1821. Van Damme, J. and Van Snick, J. (1987) Induction of hybridoma growth factor (HGF), identical to IL-6, in human fibroblasts by IL-I: use of HGF activity in specific and sensitive biological assays for IL-1 and IL-6. Dev. Biol. Stand. 69, 31. Vandenabeele, P., Jayaram, B., Devos, R., Shaw, A. and Fiers, W. (1988) Interleukin l a acts as an autocrine growth factor for RPMI 1788, an Epstein-Barr virus-transformed human B-cell line. Eur. J. Immunol. 18, 1027. Vandenabeele, P., Guisez, Y., Declercq, W., Bauw, G., Vandekerckhove, J. and Fiers, W. (1990) Response of murine cell fines to an IL-1/IL-2-induced factor in a rat/mouse T hybridoma (PC60). Differential induction of cytokines by human IL-la and IL-lfl and partial amino acid sequence of rat GM-CSF. Lymphokine Res. 9, 381. Van Snick, J., Cayphas, S., Vink, A., Uyttenhove, C., Coulie, P.G., Rubira, M.R. and Simpson, R.J. (1986) Purification and NH2-terminal amino acid sequence of a T-cell-derived lymphokine with growth factor activity for B-cell hybridomas. Proc. Natl. Acad. Sci. U.S.A. 83, 9679. Yokoi, T., Miyawaki, T., Kato, K., Kasahara, Y. and Taniguchi, N. (1990) Epstein-Barr virus-immortalized B cells produce IL-6 as an autocrine growth factor. Immunology 70, 100.