Functional characterization of cytokine autoantibodies in chronic renal failure patients

Kidney International, Vol. 45 (1994), pp. 1484—1488

Functional characterization of cytokine autoantibodies in chronic renal failure patients GERE SUNDER-PLASSMANN, STELI0 KAPI0TIs, CHRISTOPH GASCHE, and URSULA KLAAR, with

the technical assistance of BRIGITTE MAI Universitätsklinik für Innere Medizin III, Klinische Abteilung fur Nephrologie und Dialyse; Klinisches Institut für Medizinische und Chemische Labordiagnostik, Abteilung für Hamatologie und Hamostaseologie; and Universitätsklinik für Innere Medizin IV, Klinische Abteilung für Gastroenterologie und Hepatologie; Universität Wien, Wien, Austria

Functional characterization of cytokine autoantibodies in chronic renal failure patients. We have recently reported on an increase of IL-i a and IL-6 autoantibodies in patients maintained on chronic hemodialysis

treatment. The aim of the present study was to evaluate functional properties of these autoantibodies. Serum samples of more then 500 chronic renal failure patients with and without replacement therapy were screened for the presence of IL-la and IL-6 autoantibodies by second antibody precipitation. The neutralizing capacity of IL-la autoantibody serum was studied by immunofluorescence flow cytome-

try analysis of IL-la induced expression of E-selectin (ELAM-l, CD62e) on HUVEC (human umbilical vein endothelial cells). Results of

hemodialysis patients [6—16], these autoantibodies might have

some influence on cytokine function in this clinical setting. Furthermore, cytokine autoantibodies could be detected in patients suffering from various other diseases such as septicemia [1, 17], rheumatoid arthritis [181, skin disorders [19, 20], and even in some healthy subjects [17, 21, 22]. Since the pathophysiological significance of cytokine autoantibodies is a matter of debate, it was the aim of our study to evaluate the functional properties of these autoantibodies.

these inhibition studies were confirmed with IgG preparations from antibody positive sera, purified by affinity chromatography. Functional studies on IL-6-dependent B9 cell proliferation were performed with IL-6 autoantibody positive sera, and quantitated with the colorimetnc MTT assay, IL-la induced expression of E-selectin on HUVEC (considered 100% positive cells) was inhibited by each IL-la autoantibody

positive serum sample (N = 13; anti-IL-la activity: 7.62 to 57.52% binding). Inhibition of E-selectin expression by IL-Ia autoantibodies ranged from 0.11 to 80.22% positive cells (0.15 to 92.31% mean fluorescence intensity). A strong correlation of E-selectin expression with IL-la autoantibody concentration was observed (P < 0.005). Furthermore, IgG eluates from autoantibody positive patients inhibited E-selectin expression to 41.0 23.1% positive cells if compared with 83.7 5.7% positive cells of the IgG depleted serum samples. IgG eluates and IgG depleted sera derived from anti-IL-la negative serum samples had no major influence on E-selectin expression (79.9 28.5% and 77.9 16.8%, respectively). IL-6 autoantibody sera (N = 11; anti-IL-6 activity: 5.85 to 53.28% binding) had no inhibitory effect on B9 7.1%, 1:100: 106.8 9.2%, 1:10: cell proliferation (1:1000: 102.1 146.2 26.6% proliferation). We conclude that IL-la autoantibodies in

chronic renal failure patients deliver inhibitory activity on IL-la, at least in vitro. IL-6 autoantibodies were not found to influence IL-6 dependent proliferation of B9 cells.

Recent work provides compelling evidence that naturallyoccurring cytokine antagonists are involved in regulatory mechanisms of the immune system [1—31. We have recently reported

on an increase of IL-i a and IL-6 autoantibodies in serum of 32.5% and 18.2% of hemodialysis patients, respectively [4, 5]. As the activity of IL-i and IL-6 is supposed to be increased in

Methods

Cytokine autoantibodies Serum samples of more than 500 chronic renal failure patients with and without replacement therapy were screened for the presence of IL- 1 a and IL-6 autoantibodies as previously described [4, 5, 18]. Following the addition of either 1251-IL-la or '251-IL-6, autoantibody-bound '251-IL-1a11251-IL-6 was separated by a second antibody immunoprecipitation followed by

centrifugation. Pellets were counted and results denoted as percentage IgG-bound '251-IL-laI'251-IL-6 of total '251-IL-la/ '251..IL..6 activity. Reagents were from BUhlmann Lab AG (Basel, Switzerland), and '251-IL-6 was obtained from Amersham (Buckinghamshire, England). The cut off value for the normal range of IL-la autoantibodies was 3.68% binding and for IL-6 autoantibodies 2.87% binding [5].

Preparation of IgG eluates and IgG depleted sera Serum samples of patients with high IL-la autoantibody activity and autoantibody negative serum samples of healthy subjects were loaded on a HiTrap Protein G affinity chromatography column (Pharmacia, Belgium). Bound IgG was eluted with 0.1 M glycine HC1, the pH of the eluted fractions was adjusted with 1 M Iris. Purity of eluted IgG was proven by SDS-PAGE (Phast-Gel, Pharmacia, Belgium) and protein concentration was determined using the bicinchoninic acid method (Micro BCA Protein Assay Reagent Kit, Pierce, USA). Cell cultures Cell cultures of HUVEC (human umbilical vein endothelial

Received for publication April 20, 1993 and in revised form December 23, 1993 Accepted for publication December 23, 1993

cells) were performed according to Jaffe et al as described previously [23, 24]. Umbilical cord endothelial cells were

© 1994 by the International Society of Nephrology

pooled and seeded on 35 mm culture wells (Costar, Cambridge,

1484

1485

Sunder-Flassmann et al: Characterization of cytokine autoantibodies

Table 1. Influence of IL-la autoantibody sera on E-selectin Massachusetts, USA) precoated with 1 mglcm2 fibronectin expression on HUVEC (Sigma Chemical Co., St. Louis, Missouri, USA). Cells were grown in RPMI-1640 containing 1% L-glutamine, 100 lU/mi Patient IL-la autoantibody no. % Positive cells % mfi % binding penicillin G, 100 mg/ml streptomycin, 5 mg/rn! fungizone, supplemented with 5% fetal calf serum (FCS). Reaching con1 57.52 0.1! 0.15

fluence within three to seven days, the average cell number per well was about 5.5 0.7 x iO cells. Stimulation of E-selectin

(ELAM-l, CD62e) expression was achieved by incubating HUVEC monolayers for four hours with rhulL-la at 10 U/ml (5 x 108 U/mg; British Biotechnology, Oxford, UK) in 1 ml medium [25]. B9 cells (provided by M. Schwabe, NC!, FCRDC, Frederik, Maryland, USA) were maintained in RPMI-l640, supplemented

with 1% L-glutamine, penicillin G (100 lU/mi), streptomycin (100 mg/mi), 10% heat-inactivated FCS and rhulL-6 (British Biotechnology) at 50 pg/dl [26]. Culture media were from Gibco (Grand Island, New York, USA). Inhibition studies on IL-I a induced expression of E-selectin on HUVEC

2 3 4 5 6 7 8 9

10 11

12 13

14 15

16 17 18

19

20

44.13 37.58 35.18 31.36 23.00 18.09 17.29 12.34 10.36 7.94 7.89 7.62 1.06 0.13 0.11 0.05 0.98 0.96 0.77 0.76

5.93 14.60

38.90 64.83 40.82 5.39 25.88 58.88 29.74 58.40 80.22 61.72 82.49 78.56 38.23 99.54 90.28 101.76 99.40 97.71

2.38 18.57 41.69 75.63 45.71 0.10 27.62 92.31 40.04 44.05 81,55 49.17 66.90 83.57 34.91 98.31 81.76 115.37 127.45 89.87

Neutralizing IL-la activity was tested by incubating IL-la 21 autoantibodies containing sera (final dilutions with medium Patients 1 through 17 have chronic renal failure; 18 to 21 are healthy 1:20, and in some experiments 1:200 and 1:500) either with subjects. Abbreviation mfl is mean fluorescence intensity. rhulL-la (10 U/mi) or with medium alone as control for one hour at room temperature prior to coincubation with HUVEC. The same protocol was run with IgG depleted serum samples 100-, and 1000-fold dilution, prior to the proliferation assay. (final dilution 1:20) and with IgG eluates at a protein concen- One unit of IL-6 was defined as the concentration leading to half tration of 1 mg/mi. To confirm specificity, E-selectin expression was inhibited by a monoclonal mouse anti-IL-i a antibody (data

not shown). Control experiments were performed with IL-i a autoantibody negative uremic and normal donor serum samples in 20-fold dilutions and with medium alone, as described above.

Cell viability of HUVEC previous to staining was more than

maximal B9 cell proliferation (25 to 250 pg/mi). IL-6 specificity of cell proliferation was confirmed by inhibition studies using a mouse anti-IL-6 mAb. Proliferation of B9 cells was determined with the MiT assay [27]. MTT [3- (4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazo-

hum bromide, Sigma Chemical Co.] was dissolved in PBS (5

90% as determined by Trypan blue exclusion. After mild mg/mi = stock solution) and sterile filtered. This stock solution collagenase detachment and two further washing steps, analysis was diluted 1:10 with medium and 25 j.d were added to each well of surface molecule expression on HUVEC was performed by after discarding 125 ml of the supernatant. Plates were incuflow cytometry analysis (FACScan, Becton Dickinson, Moun- bated at 37°C. After four hours, 125 pJ of 0.04 N HC1 in tain View, California, USA). Fifty microliter cell suspension isopropanol were added to each well and plates were shaken to

(S x i05 HUVEC/ml) was incubated with 50 ml of anti-E- dissolve MIT crystals. Within at least one hour, plates were selectin mAb (4Db, Immunotech, Marseille, France; 30 mm at room temperature), followed by staining with fluorescein labeled goat f(ab')2 anti-mouse IgG + 1gM antibody (30 mm at 4°C), fixation in 0.5% paraformaldehyde, and analyzing using the LYSIS II software (Becton Dickinson).

read on a micro-ELISA reader at a wavelength of 570 nm and a

reference wavelength of 620 nm. Results of the inhibition studies are presented as percentage of IL-6 induced proliferation of B9 cells at the optimal IL-6 concentration. Statistical analysis

Inhibition studies on IL-6 dependent proliferation of B9 cells For the set up of the inhibition studies, B9 cells were washed twice (in PBS, pH 7.2) to eliminate residual amounts of IL-6.

Viability (>90% in all experiments) was determined using

Paired and unpaired data were calculated with Student's t-test or Pearson's correlation coefficient; P < 0.05 was considered to be significant.

Results Trypan blue exclusion. One hundred microliter cell suspensions IL-i a and IL-6 autoantibody positive sera per well (4 x iO B9 cells/p.l culture medium) were transferred to a 96-well flat-bottomed culture plate (Costar) and incubated Serum samples from 13 IL-la and 11 IL-6 autoantibody with 100 p.I rhulL-6 (1 ng/ml) in serial 1:2 dilutions in medium as positive uremic donors, as well as antibody negative uremic and positive control for 96 hours in quadruplicates (humidified air, normal donors' serum samples were selected for the inhibition 37°C, 5% C02). Inhibition of IL-6 activity with IL-6 autoanti- studies. IL-la and IL-6 autoantibody activities of the individual body containing serum IL-6 autoantibody positive (N = 11) as serum samples are indicated in Tables 1 and 2. Three further serum samples of IL-la autoantibody positive well as autoantibody negative uremic (N = 3) and normal donors (N = 3) sera were incubated with rhulL-6, or medium patients and of three autoantibody negative healthy subjects alone as control, for one hour at room temperature at a final 10-, were depleted from IgG by affinity chromatography. Anti-IL-la

1486

Sunder-Plassmann et a!: Characterization of cytokine autoantibodies

45

Table 2. Influence of IL-6 autoantibody sera on B9 cell proliferation Patient no. 1

2 3 4 5 6 7 8 9 10

Proliferation % OD

IL-6 autoantibody

%binding

1:1000

1:100

1:100

53.28 29.64 17.55 15.25 12.84 7.27 7.18 7.14 7.07 6.10 5.85 0.81

101.1 117.1 106.3

97.3 119.0 117.1 120.4 101.8 109.4

125.5 113.3 108.2 192.3 148.1 162.3 119.3 146.7 162.5 159.0

0.10

97.5 95.1

95.8 101.8 105.4

88.4 105.8 102.5

98.4 100.9

95.0 97.2 109.8

107.4 100.6 115.4 102.1 207.7

0) 30 a)

C a)

0 a)

a-

15

0

170.6 280.0 112.6 209.2 95.5 106.1 406.2

1/200 1/500 Medium IL-i 1/20 Fig. 1. Inhibition of E-selectin expression on HUVEC by serial dilu-

Patients 1 through 14 have chronic renal failure; 15 to 17 are healthy subjects.

10.32% mfi). Increasing dilutions of anti-IL-la serum up to 1:500 hardly inhibited E-selectin expression (1:200: 34.82% positive cells, 26.77% mfi; 1:500: 39.93% positive cells, 33.58% mfi).

11

12 13 14 15 16 17

0.10

0.16 0.10 0.01

106.2 100.2

98.6

92.4 87.7

108.0 135.4

tions of IL-la autoantibody serum in one patient. Symbols are: () positive cells; () mfi. In this experiment 43.04% of HUVEC expressed E-selectin following stimulation with IL-la (32.63% mfi). Incubation with IL-la autoantibody serum (10.36% binding) resulted in a nearly

complete inhibition of E-selectin expression (8.93% positive cells,

activity of the antibody positive sera was 17.28, 36.21, and 57.82%, respectively. IL-i a autoantibody activity of the three normal donor sera was <3.68% in each case.

60

Neutralizing activity of IL-i a autoantibody serum Flow cytometry analysis of HUVEC was performed to quan-

40

titate anti-IL-la mediated inhibition of IL-la induced surface expression of E-selectin. Following treatment with IL-la (10 UIml; 4 hr), cell populations weakly positive for E-selectin gained a strong positive staining pattern. Percentage of Eselectin positive cells and % mean fluorescence intensity (mfi) was calculated as follows: (sample — control): (IL-la — control) x 100.

Incubation of IL-i a with IL-i a autoantibody positive serum,

.

.

.

>-.

.

>

( 0

20

•

-J

. •

C

.

0

0

35

: .

••

70

105

E-selectin expression, % positive cells

performed prior to coincubation with HUVEC, blocked ECoincubation of HUVEC with IL-la autoantibody serum selectin expression strictly depending on antibody concentra- selectin. resulted in an inhibition of IL-la induced expression of E-selectin. The Fig. 2. Correlation of IL-I a autoantibody activity with expression of E-

tion and final dilution of serum samples (N = 13; 0.11 to 80.22% positive cells, 0.10 to 92.31% mfi). Individual data are shown in Table 1.

concentration of IL-la autoantibodies in individual serum samples (N = 21) correlated significantly with the decrease of E-selectin expression (Y = —0.385x + 36.541, r = 0.786, P

< 0.005).

Serial serum dilutions up to 1:500 from one patient with 10.36% antibody activity abolished the inhibitory effect of IL-la autoantibodies on E-seiectin expression completely (Fig. 1). In further experiments, normal (N 4) and uremic (N = 4)

induced expression of E-selectin in nearly the same manner as 13.5% 23.1% and 30.3 untreated serum samples (41.01 donors' sera which were negative for IL-la autoantibodies positive cells, respectively), whereas the IgG depleted serum showed no major influence on E-selectin expression in all but had no major influence on E-selectin expression (83.7 5.7% one of the serum samples (90.28 to 101.76% positive cells, 81.76 positive cells). In anti-IL-la negative serum samples the correto 127.45% mfi, and 38.23 to 99.54% positive cells, 66.9 to sponding results were 79.9 28.5, 77.9 16.8, and 84.4 98.31% mfi, respectively). The correlation of anti-IL-i a activity 20.1% positive cells, respectively (Fig. 3). and autoantibody mediated inhibition of E-selectin expression Neutralizing activity of IL-6 autoantibodies was highly significant (N = 21, r 0.786, P < 0.005; Fig. 2). Using the colorimetric MTT-assay, IL-6 dependent proliferNeutralizing activity of IgG depleted serum and purified IgG ation of B9 cells was evaluated. Maximal proliferation was IgG eluates from three IL- 1 a autoantibody positive and three observed at IL-6 concentrations of 25 to 250 pg/mi, and the negative serum samples, isolated by affinity chromatography, respective ODs in individual experiments were assigned as were adjusted to a protein concentration of 1 mg/mI. IgG eluates 100% proliferation (Table 2). Preincubation of IL-6 with serial from autoantibody positive serum samples inhibited IL-la dilutions of anti-IL-6 autoantibody serum samples did not result

Sunder-Plassmann et a!: Characterization of cytokine autoantibodies

120

amounts of unlabeled IL-6 completely displaced '251-IL-6 from anti-IL-6 IgG [22].

0 C', C')

ci)

xa) C ci) ci)

40

ci,

w 0

1487

IiIl1

Serum Ig depleted serum lg eluates Fig. 3. Inhibition studies of E-selectin expression on HUVEC with IgG eluates and IgG depleted sera of IL-la autoantibody positive () and negative () sera. IL-la induced expression of E-selectin on HUVEC

Whether human cytokine autoantibodies have inhibitory functions or facilitate transport and binding to specific target sites in vivo, by protecting cytokines from urinary excretion, proteolytic degradation or binding to other receptors, remains to be elucidated. Bendtzen et al assumed that cytokine autoantibodies might be encoded in germ-line genes, and get switched

on upon demand [1]. This assumption is strenghtened by the observation of a very rapid occurrence of IL-i a and TNFa autoantibodies in patients suffering from septicemia [1, 17].

In our study, IL-la induced expression of E-selectin on

HUVEC could nearly completely be blocked by uremic serum containing IL-ia autoantibodies in a dose dependent manner. Specificity of this phenomenon was proven by experiments is inhibited by crude serum and IgG eluates of IL-la autoantibody using IgG depleted serum as well as IgG eluates, isolated by positive sera, but not by IgG depleted positive sera or IL-la autoantiaffinity chromatography. On the other hand, anti-IL-i a negabody negative serum samples, IgG eluates, or IgG depleted sera (N = tive uremic and normal donors' sera and IgG eluates failed to 3 in each experiment). influence IL-i a induced expression of E-selectin on HUVEC in most cases. These autoantibodies might protect some hemodialysis patients from proinflammatory effects of IL-i, such as the in an autoantibody dependent abrogation of B9 cell proliferation increase of leukocyte adherence to the vessel wall. (N = 11; 1:1000: 102.1 7.1%; 1:100: 106.8 9.2% proliferaIL-6 autoantibodies in chronic renal failure patients sera did tion), as it was seen with the mouse IL-6 mAb. However, not inhibit cytokine function in vitro. However, the failure of incubation with a ten-fold serum dilution resulted in an unspe- IL-6 autoantibodies to block IL-6 dependent proliferation of B9 cific stimulation of B9 cell proliferation (146.2 26.6%, P < cells might be due to the presence of multimeric IL-6 molecules, 0.05). IL-6 autoantibody negative normal (N = 3) and uremic (N not completely neutralized by IL-6 autoantibodies [29]. Fur= 3) donors' sera had no inhibitory activity on B9 cell prolifer- thermore, IL-6 autoantibodies might bind to IL-6 epitopes ation (1:1000, 93.4 6.3% and 99.6 5.8%; 1:100, 114.0 distinct from the receptor binding part of the molecule. Addi19.1% and 141.7 57.6%, respectively), but stimulated prolif- tionally, the influence of different cytokine binding molecules, eration at a tenfold dilution (202.6 176.4% and 200.6 80.0%, such as soluble receptors or receptor antagonists on these assay respectively) in some cases. systems, cannot be ruled out [2, 301. Applying higher concentrations of anti-IL-6 positive and negative sera resulted in a Discussion more pronounced proliferation of B9 cells in some experiments. IL-la and IL-6 autoantibodies can be detected with high This was not unexpected and might, at least in part, be due to prevalence in hemodialysis patients, in contrast to other renal an increased protein concentration, or a synergistic effect of failure patients or normal individuals [4, 5]. In the present other cytokines, as it was recently demonstrated for IL-2 [31]. study, we focused our interest on putative functional properties In summary, this study demonstrates the presence of a factor of these autoantibodies. Our results clearly demonstrate that in uremic serum, interfereing with the endothelial "adhesion IL-i a autoantibodies, present in uremic serum, block the IL-i a cascade" [32]. Our data confirm a neutralizing capacity of mediated expression of E-selectin on HUVEC, whereas IL-6 IL-ia, but not of IL-6 autoantibodies, in chronic renal failure autoantibody sera failed to specifically inhibit IL-6 activity, at patients maintained on hemodialysis treatment. least IL-6 induced proliferation of B9 cells. Concerning other in vitro studies, some authors were able to Reprint requests to Dr. Gere Sunder-Plassmann, Klinische Abteilung demonstrate an inhibitory effect of IL-i a autoantibodies on für Nephrologie und Dialyse, Universitätsklinik für Innere Medizin III, IL-i a binding to its receptor on endothelial, synovial and T Ailgemeines Krankenhaus der Stadt Wien, Universität Wien, Wahring-

cells, and on IL-i a dependent proliferation of thymocytes [1, 18, 21]. Furthermore, endothelial adherence of neutrophils could be blocked by anti-IL-i a IgG containing serum [28]. Additionally, tissue distribution and the rapid clearance of exogeneous IL-la in rodents was heavily influenced by IL-la autoantibodies [19, 201.

Little information exists regarding IL-6 autoantibodies. Hansen described interference of polyclonal anti-IL-6 IgG, derived from normal donors' sera, with recovery of exogeneous

IL-6 in an ELISA system. As demonstrated by gel filtration

ergurtel 18-20, A-1090 Wien, Austria.

References 1. BENDTZEN K, SVENSON M, JONSSON V, HIPPE E: Autoantibodies to cytokines—Friends or foes? Immunol Today 11:167—169, 1990

2. DINARELLO CA, THOMPSON RC: Blocking IL-l: Interleukin I receptor antagonist in vivo and in vitro. Immunol Today 12:404— 410, 1991

3. LARRICK JW: Native interleukin-i inhibitors. Immunol Today 10: 61—66, 1989 4. SUNDER-PLASSMANN G, SEDLACEK

PL, SUNDER-PLASSMANN R, DERFLER K, SWOBODA K, FABRIzII V, HIRSCHL MM, BALCKE P:

techniques, these antibodies coprecipitated 1251-IL-6, and were shown to bind predominantly to monomeric IL-6, and second antibody precipitation indicated that more than one IgG mole-

ney mt 40:787—791, 1991 5. SEDLACEK PL, SUNDER-PLASSMANN G, KLAAR U, DERFLER K,

cule bound to a single IL-6 molecule. Additionally, excess

SWOBODA K, DRUML W, K0vARIK J: Anti-IL-la and anti-IL-6

Anti-interleukin-1 a autoantibodies in hemodialysis patients. Kid-

1488

Sunder-Plassmann et a!: Characterization of cytokine autoantibodies

autoantibody activity in chronic renal failure patients. (abstract), J Am Soc Nephrol 3:665, 1992 6. HENDERSON LW, KOCH KM, DINARELLO CA, SHALDON S: He-

modialysis hypotension: The interleukin hypothesis. Blood Purf 1:3—8, 1983

7. BINGEL M, LONNEMANN G, SHALD0N S, KOCH KM: Human

interleukin-1 production during hemodialysis: The influence of dialysis membranes. Nephron 43:161—163, 1986 8. LONNEMANN G, BINGEL M, FLOEGE J, KOCH KM, SHALDON S, DINARELLO CA: Detection of endotoxin-like interleukin-1-inducing activity during in vitro dialysis. Kidney mt 33:29—35, 1988 9. HAEFFNER-CAVAILLON N, CAVAILLON JM, CIANCIONI C, BACLE F, DELONS S, KAZATCHKINE MD: In vivo induction of interleukin-1

during hemodialysis. Kidney mt 35: 1212—1218, 1989 10. HERBELIN A, NGUYEN AT, ZINGRAFF J, URENA P, DESCAMPSLATSCHA B: Influence of uremia and hemodialysis on circulating

interleukin-1 and tumor necrosis factor a. Kidney mt 37:116—125, 1990 11. LAUDE-SHARP M, CAROFF M, SIMARD L, PUSINERI C, KAZATCH-

KINE MD, HAEFFNER-CAVAILLON N: Induction of IL-I during hemodialysis: Transmembrane passage of intact endotoxins (LPS). Kidney mt 38:1089—1904, 1990

12. SCHINDLER R, LONNEMANN G, SHALDON 5, KOCH KM, DINARELLO CA: Transcription, not synthesis, of interleukin-1 and tumor necrosis factor by complement. Kidney mt 37:85—93, 1990 13. HERBELIN A, URENA P, NGUYEN AT, ZINGRAFF J, DESCAMPS-

LATSCHA B: Elevated circulating levels of interleukin-6 in patients with chronic renal failure. Kidney ml 39:954—960, 1991 14. NAKAHAMA H, TANAKA Y, SHIRAI D, MIYAZAKI M, IMAI N, YOKOKAWA T, OKADA M, KUBORI S: Plasma IL-6 levels in

continuous ambulatory peritoneal dialysis and hemodialysis patients. Nephron 61:132—134, 1992 15. CHEUNG AK: Biocompatibility of hemodialysis membranes. JAm Soc Nephrol 1:150—161, 1990 16. DINARELLO CA: Cytokines: Agents provocateurs in hemodialysis? Kidney mt 41:683—694, 1992 17. FOMSGAARD A, SVENSON M, BENDTZEN K: Auto-antibodies to

20. SAURAT JH, SCHIFFERLI J, STEIGER 0, DAYER JM, DIDIERIEAN L:

Anti-IL-la antibodies in humans: Characterization, isotype distri-

bution, and receptor-binding inhibition—Higher frequency in Schnitzlers syndrome. J Allergy Clin Immunol 88:244—256, 1991 21. SVENSON M, POULSEN LK, FOMSGAARD A, BENDTZEN K: IgG

autoantibodies against interleukin-la in sera of normal individuals. Scand J Immunol 29:489—492, 1989 22. HANSEN MB, SVENSON M, DIAMANT M, BENDTZEN K: Anti IL-6

antibodies in normal human serum. Scand J Immunol 33:771—781, 1991

23. JAFFE EA, NACHMANN RL, BECKER CG, MINIcK CR: Culture of human endothelial cells derived from umbilical veins. Identification

by morphologic and immunologic criteria. J Clin Invest 52:2745— 2752, 1973 24. KAPI0TI5 S, BESEMER J, BEVEC D, VALENT P, BETTELHEIM P,

LECHNER K, SPEISER W: Interleukin-4 counteracts pyrogen-induced downregulation of thrombomodulin in cultured human vascular endothelial cells. Blood 78:410—415, 1991 25. Lu5CIN5KAs FW, BROCK AF, ARNAOUT MA, GIMBRONE MA: Endothelial-leucocyte adhesion molecule-l-dependent and leucocyte (CD! 1/CD 18)-dependent mechanisms contribute to polymorphnuclear leucocyte adhesion to cytokine-activated human vascular endothelium. J Immuno! 142:2257—2263, 1989 26. AARDEN LA, DE GROOT ER, SCHAAP OL, LANDSDORP PM: Production of hybridoma growth factor by human monocytes. Eur J Immunol 17:1411—1416,1987

27. MOSMANN T: Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays. J Immuno! Met/i 65:55—63, 1983 28. SUZUKI H, AKAMA T, OKANE M, KoNo I, MATSUI Y, YAMANE K,

KA5HIWAGI H: Interleukin-l-inhibitory IgG in sera from some patients with rheumatoid arthritis. (abstract) Arthritis Rheum 32: 1528, 1989 29. WIJDENES J, CLEMENT C, KLEIN B, MOREL-FOURRIER B, VITA N,

FERRARA P, PETERS A: Human recombinant dimenc IL-6 binds to

its receptor as detected by anti-IL-6 monoclonal antibodies. Mo!

TNFa in healthy humans and patients with inflammatory diseases and gram-negative bacterial infections. Scand J Immunol 30:2 19—

Immunol 28:1183—1192, 1991 30. HONDA M, YAMAMOTO S, CHENG M, YASUKAWA K, SuzuKI H, SAIT0 T, OsuGi Y, TOKUNAGA T, KEsHIM0T0 T: Human soluble

223, 1989 18. SUZUKI H, KAMIMURA J, AYABE T, KA5HIwAGI H: Demonstration

IL-6 receptor: Its detection and enhanced release by HIV infection. Jlmmunol 148:2175—2180, 1992

of neutralizing autoantibodies against IL-la in sera from patients

31. CHIRMULE N, OYAIZU N, KALYANARAMAN VS, PAHWA S: Misin-

with rheumatoid arthritis. J Immuno! 145:2140—2146, 1990 19. DIDIERJAN L, SCHIFFERLI J, STEIGER 0, SAURAT JH: IgG autoanti-

bodies to IL-Ialpha in the serum of patients with skin diseases. Incidence, characterization and effect on the blood clearence of IL-lalpha. (abstract) J Invest Dermato! 94:575, 1990

terpretation of results of cytokine bioassays. J Immunol Met/i 137:141—144, 1991

32. PARDI R, INVERARDI L, BENDER JR: Regulatory mechanisms in leukocyte adhesion: Flexible receptors for sophisticated travellers. Immuno! Today 13:224—230, 1992