In vivo T cell preactivation in chronic uremic hemodialyzed and non-hemodialyzed patients

Kidney International, Vol. 36 (1989), pp. 636—644

CLINICAL INVESTIGATION

In vivo T cell preactivation in chronic uremic hemodialyzed and non-hemodialyzed patients GENEVIEVE BEAURAIN, CATHERINE NARET, LuIsA MARCON, GILLES GRATEAU, TILMAN DRUEKE, PABLO URENA, DAVID L. NELSON, JEAN-FRANcOIS BACH, and LUCIENNE CHATENOUD Inserm U 25, CNRS UA 122, Ass. Cl. Bernard Hôpital Necker and Centre dHémodialyse Edouard Rist, Paris, France; Metabolism Branch, National Cancer Institute, National Institute of Health, Bethesda, Maryland, USA: Département de Néphrologie, INSERM U90 and Centre d'Hémodialyse, Hôpital Necker, Paris, France

In vivo T cell preactivation in chronic uremic hemodialyzed and

apy, notably hemodialysis, the immunity of uremic hemodialyzed patients remains poor [15]. The biological in vitro counpatients undergoing regular hemodialysis treatment and 20 uremic terpart of this clinically evident immunodeflciency is the low patients prior to the onset of renal replacement therapy. In hemodia- capacity of purified lymphocytes from uremic patients to respond to antigens, mitogens and monoclonal antibodies triggerlyzed patients, abnormally increased proportions of circulating T cells non-hemodialyzed patients. The production and targeting of a major T cell derived lymphokine, Interleukin 2 (IL-2), were studied in 23 uremic

spontaneously expressing high affinity IL-2 receptors (IL-2 Rec) were detected; they bound a monoclonal antibody specifically directed to the IL-2 Rec 55 kDa chain (Tac antigen) (mean SEM: 7.12 0.81% in patients vs. 2.15 0.39% in normal controls, P < 0.0001) and significantly proliferated in presence of human recombinant IL-2 alone (mean sEM: 5438 729 cpm in patients vs. 1647 244 cpm in normal controls). Hemodialyzed patients also exhibited significantly increased serum levels of soluble IL-2 receptor (mean SEM: 4036 947 ti/mI in patients vs. 253 29 U/mI in normal controls, P <0.001). Moreover, a significantly decreased IL-2 activity was detected in the supernatants of stimulated T cells from hemodialyzed patients (mean SEM: 0.93 0.12 U/mI in patients vs. 2.49 0.22 U/mI in normal controls, P < 0.0001). In nine hemodialyzed patients who were analyzed before and immediately after the hemodialysis session no acute modifications of

the various parameters analyzed were detected. Although less profound, a similar pattern of T cell abnormalities was observed in the uremic non-hemodialyzed patients studied. These data suggest that the deficient IL-2 activity in uremic patients may be at least in part related to increased absorption by IL-2 receptor present in abnormally high amounts on T cells and in the serum. Furthermore, they underline that although exacerbated by hemodialysis treatment, a latent state of T cell preactivation is associated with end-stage chronic renal failure.

Chronic renal failure induces a state of immunodeficiency

ing the T3-Ti T cell receptor complex [1—3, 16—18].

Several hypotheses have been put forward to explain the immunosuppression associated with chronic uremia, including an intrinsic defect of uremic effector T lymphocytes [19] and/or the presence of an abnormal immunoregulation mediated by regulatory T lymphocytes (such as, helper and suppressor T cells), monocytes or soluble serum factors [20—26]. In a previous report we have shown that uremic hemodialyzed patients present an abnormal bioavailability and targeting of one major T cell derived soluble factor, namely, interleukin

2 (IL-2) [27], that plays an essential role in cell-mediated immunity by promoting cell proliferation and differentiation [281. The preliminary study showed a decreased IL-2 activity in culture supernatants from stimulated hemodialyzed patients' T cells [27]. Moreover, uremic hemodialyzed patients presented high levels of peripheral T cells showing phenotypic and func-

tional signs of preactivation [27]. This particular subset of T cells spontaneously expressed the interleukin 2 receptor (IL-2 Rec) at their cytoplasmic membrane [27]. The expression of IL-2 Rec is considered as a marker of activation since it is absent from normal resting T cells and exclusively appears following antigenic or mitogenic stimulation [29]. It was sug-

that mainly involves T cell-mediated immune responses [1—3]. gested that these activated T cells could participate to the Studies show cutaneous anergy [4], prolonged allograft survival defective IL-2 activity observed in hemodialyzed patients by [5, 6], a decreased antibody response to thymus dependent promoting increased absorption of the lymphokine. antigens like vaccination with HB virus [7—10], influenza virus In the present report our previous results were confirmed and [11], as well as an increased incidence of infections not related extended to a larger series of hemodialyzed patients. Moreover,

to the dialysis treatment such as mycobacterial infections a group of uremic non-hemodialyzed patients was studied in [12—14]. In spite of progressive improvement over the past few parallel, to determine whether the uremic state per se or years of the various methods used for renal replacement ther-

Received for publication September 28, 1988 and in revised form March 9, 1989 Accepted for publication April 27, 1989

© 1989 by the International Society of Nephrology

changes induced by the hemodialysis procedure could be responsible for the T cell preactivation observed. Methods

Patients Forty-three patients presenting end-stage chronic renal failure were included in the study.

636

Beaurain et a!: In vivo T cell activation and uremia

Twenty-three of these patients were stabilized on renal replacement therapy for more than six months prior to the

637

Immunofluorescence test Monoclonal antibodies used for labeling assays were OKT3

study. Patients underwent hemodialysis twice a week for four (IgG2a), OKT4 (IgG2b), OKT8 (IgG2a) (Ortho Pharmaceutical

to five hours using Cuprophan (Hospal, Lyon, France) or Corporation, Raritan, New Jersey, USA) that respectively polyacrylonitrile membranes and water pretreated by filtration. recognize total human T cells (CD3), helper/inducer (CD4) This group of patients included 16 men and 7 women whose ages ranged from 15 to 74 years old (mean 5EM: 38.7 4.0 years). In these hemodialyzed patients, chronic renal failure was secondary to chronic glomerulonephritis (9 patients), Alport's disease (1 patient), nephroangiosclerosis (3 patients), chronic instertitial nephritis (3 patients), nephronophitisis (I

and cytotoxic/suppressor T cells (CD8) [30]. In addition, Tact T cells were detected using lOT l4a (Immunotech, Marseille, France), a rat monoclonal antibody that recognizes the human Tac antigen (that is, the IL-2 receptor). This antibody was used with OKT3 in double labeling experiments. Indirect immunofluorescence was performed as previously

patient), congenital malformations (3 patients), diabetic glomerulopathy (1 patient), polycystic kidney disease (I patient) and

described [311 on freshly stimulated E cells or following

renal lithiasis (1 patient). The other 20 patients presented end-stage renal failure (creatininemia ranging from 460 to 1440 /LM/liter) and were studied before the institution of hemodialy-

sis. This group included 11 men and 9 women, aged 20 to 89 years (mean SEM 55.5 4.8 years). In these patients chronic renal failure was secondary to chronic glomerulonephritis (5 patients), nephroangiosclerosis (2 patients), congenital malformation (6 patients), chronic insterstitial nephritis (5 patients) and polycystic kidney disease (2 patients). Patients in whom chronic renal failure was secondary to an immunologic disease were excluded. None of the patients received prior to or during the study any treatment known to interfere with the immune system (steroids or other immunosuppressants, /3 blockers, etc.). At the time of the study none of the patients presented clinical or biological signs of an evolving infectious process. All were free of human immunodeficiency virus (HIV) infection.

Concerning hepatitis B (HB) virus infection, one of the

incubation with recombinant IL-2. Briefly, the cells (10 to 20 x 106/ml) were incubated for 30 minutes at 4°C with appropriate dilutions of OKT3, OKT4 and OKT8. After two washes in cold HBSS containing 5% fetal calf serum (FCS) and 0.2% sodium

azide, the cells were labeled for 30 minutes at 4°C with fluorescein (FITC)-conjugated goat anti-mouse IgG antiserum (Nordic, Tilburg, The Netherlands). Following washing and resuspension, one drop was examined under a Leitz Dialux microscope equipped for epifluorescence and 200 cells per slide were counted. In double labeling experiments cells were first incubated with a mixture of OKT3 and IOTl4a. Cells were then washed and labeled with a mixture of rhodamine (TRITC)labeled goat anti-mouse IgG antiserum (Nordic, Tilburg, The Netherlands) and FITC-labeled mouse anti-rat, K light-chain

monoclonal antibody (Immunotech) to detect OKT3 and IOTl4a binding, respectively. Results were expressed as percentage values of each given T cell subset with respect to the total number of mononuclear

uremic patients not yet on dialysis was positive for HBs antigen cells counted and as % CD4/% CD8 cell ratio. (Ag), one was HBs Ag positive and also presented spontaneous Finally the percentage of doubly labeled CD3 Tack lymphoanti-HBs and anti-HBc antibodies. Two out of the 23 hemodi- cytes was expressed by reporting the number of single and

alyzed patients were HBs Ag positive and also presented

doubly labeled cells to that of all mononuclear CD3 cells

anti-HBs antibodies; 11 other patients were HBs Ag negative scored and presented spontaneous anti-HBs antibodies; the remaining % CD3 lOT 14a cells patients who were negative for HBs Ag and specific antibodies % CD3Tac cells = ____________________ x 100 underwent vaccination. % CD3 cells In the case of hemodialyzed patients blood samples were Mitogen-induced interleukin 2 production and proliferation collected immediately before dialysis. For nine hemodialyzed patients samples were collected twice, before dialysis and Culture medium was RPMI 1640 (Gibco, Paisley, Scotland, shortly after the end of the dialysis session. UK) supplemented with penicillin (100 units per ml), streptoAs a control population, 23 age- and sex-matched normal mycin (100 sg/ml), and FCS (2%). E cells were adjusted to 1 volunteers were studied. All blood samples were collected in x 106 cells per ml distributed (I mI/well) in 24-well LiMbro heparin containing vacutainers. Plasma was recovered and culture plates (Falco) and stimulated with 1 jsg of PHA-P frozen at —80°C until tested, and the cell pellet including red (Pharmacia) per ml. Supernatants were collected after 48 hours and white blood cells was used to recover the mononuclear of culture (in humidified 95% air/5% CO2 at 37°C), centrifucells. gated, sterilized by filtration (45-sm pore size), and stored at —20°C until tested for IL-2 activity. In parallel the capacity of E cells to proliferate following Cellular preparations PHA stimulation was also assessed. E cells were cultured in Peripheral blood mononuclear cells (PBM) were isolated by flat-bottomed LiMbro microplates (0.2 x l0 cells in 0.2 ml of centrifugation at 400 x g on a Ficoll-Paque (Pharmacia, Upp- complete culture medium) in presence of 0.4 tg/ml of PHA-P. sala, Sweden) gradient. PBM were recovered at the interface Proliferation was assayed after 72 hours of culture by means of and washed twice in Hanks' balanced salt solution (HBSS; a 16-hour [3H] thymidine incorporation. Results were expressed Eurobio, Paris, France). When required, PBM were separated by means of the proliferation index that was calculated as into T cells and non-T cells by rosetting with neuraminidase follows: (Behringwerke, Marburg, FRG) treated sheep erythrocytes cpm (cells + PHA) — cpm (cells — PHA) [27]. Rosetting E lymphocytes thus obtained contained 92 to 95% OKT3 cells as assessed by indirect immunofluorescence. cpm (cells — PHA)

638

Beaurain et a!: In vivo T cell activation and uremia

IL-2 titration

Soluble IL-2 receptor

Detection of soluble IL-2 receptors in plasma and culture supernatants was performed by means of an already described activity on the IL-2 dependent, murine-cytotoxic T-cell line, CTLL-2. Briefly, 4 x l0 cells were cultured in round-bottomed ELISA assay [34]. Briefly, Immulon 1 flat-bottomed 96-well microplates in the presence of serial dilutions of the test microtiter plates (Dynatech Laboratories, Alexandria, Virginia, supernatants. The proliferation was analyzed after 74 hours by USA) were coated with 150 p1 of purified anti-Tac [29] at I means of MTT staining as described elsewhere [32]. Units were g/ml in carbonate buffer, pH 9.6 or carbonate buffer alone as calculated by comparison with the proliferative activity of serial a background control, and left overnight. After washing, 100 p1 dilutions of a reference supernatant containing by definition I of each plasma sample was added to the coated and control unit per ml (that corresponds to 3.6 international NIH units and wells for a two hour incubation. Following washing 100 p1 of a is produced by 48-hr PHA stimulation of normal PBM) [33]. The 1/4000 dilution of FITC-conjugated 7G7/B6 [35] (that is, a IL-2 concentration of each supernatant was calculated by using monoclonal antibody recognizing an epitope of the IL-2 receptor distinct from that recognized by anti-Tac) in phosphatea probit program. In 22 patients soluble phase radioimmunoassay (RIA) was buffered saline containing 1% FCS was added to all wells. After performed in parallel with the biological assay. The RIA kit was an additional two hour incubation, the plates were washed and provided by Medgenix (Fleurus, Belgium). Briefly, 100 ui of the 100 p1 of a 1/1000 dilution of alkaline phosphatase-conjugated samples to be tested (in duplicate at serial dilutions) were rabbit anti-FITC were added. p-Nitrophenyl phosphate (1 mgI incubated with 100 p1 of the supplied rabbit anti-human IL-2 ml) was used as the enzyme substrate. Absorbance of the wells polyclonal antiserum for 18 to 24 hours at room temperature. was determined after 30 minutes at 405 mm using a Titertek Then 100 d of the tracer 125 I-IL-2 was added to each tube and ELISA reader (Flow Laboratories, Rockville, Maryland, incubated for four hours at room temperature. Following the USA). A reference reagent consisting of the cell-free supernaaddition of anti-rabbit gamma-globulins diluted in a buffer tant of a normal IL-2-dependent human I cell line, four days containing polyethylene glycol, cellulose and Tween 20, the after stimulation with 10% IL-2 (Cellular Products, Buffalo, samples were centrifuged 15 minutes at 1500g. The supernatant New York, USA), was used in all of these studies. The was aspirated and the radioactivity of the pellet assessed in a y undiluted supernatant was assigned a value of 1000 IL-2R U/mI, Culture supernatants were thawed and tested for their IL-2

counter (LKB, Paris, France). In this RIA two reference and the absorbance values, determined by ELISA, of serial

preparations were used to set the standard curve, namely, the dilutions of this supernatant were used to generate a reference same natural IL-2 supernatant used in the biological assay and curve. human recombinant IL-2. The bound radioactivity, that is, the Statistical analysis percentage of 125 I-IL-2 bound, was determined as follows: Statistical analysis of the results was performed by means of (cpm standard or sample) — (cpm nonspecific) B/Bo the Students 1-test for paired or unpaired data. — (cpm zero standard) — (cpm nonspecific) Results where cpm (nonspecific) equals buffer + tracer, and cpm (zero standard) is the sample containing OU/ml of IL-2. The mean IL-2 Activity radioactivity introduced/tube ranged 20,000 cpm; the cpm of the IL-2 activity was evaluated in the supernatants collected zero standard ranged to 400 cpm; the cpm of the nonspecific from PHA stimulated E rosetting T cell cultures by means of the sample ranged to 10,000 cpm. The B/Bo x 100 values for each standard point were plotted biological assay using the murine CTLL-2 line. In parallel, in as a function of the IL-2 concentration using semi-logarithmic order to confirm both the sensitivity and specificity of the paper. IL-2 concentrations in test samples were then deduced biological assay, a soluble phase RIA using human '251-IL-2 as by extrapolation to the standard curve. The sensitivity of this a tracer, was used in 20 of the uremic hemodialyzed and assay was 0.5 U/mI, that is, 0.05 U/tube (according to the non-hemodialyzed patients. Figure 1 shows that a significant international NIH standard). In our laboratory this RIA was correlation was found when comparing experimental values used in parallel to the conventional CTLL-2 assay to test 200 obtained with these two tests. At the dose of mitogen used (2 culture IL-2 containing supernatants obtained by stimulating T /Lg/ml of PHA) the mean IL-2 activity present in supernatants cells either from normal controls or patients presenting different collected from normal controls' E rosetting T cells was 2.49 pathologies. In all cases a perfect correlation between results 0.22 U/mI (mean sEM). In parallel, the cell proliferation in the same cultures used to obtained with those two tests was observed. evaluate PHA induced IL-2 activity was also studied. In normal controls the observed mean proliferation index was 184.8 Sensitivity of T cells to exogenous IL-2 22.4 (mean SEM). Uremic non-hemodialyzed patients. Mean IL-2 activity obFreshly isolated E rosetting T cells were cultured in triplicate (0.5 x 106/ml) in round-bottomed microtiter plates (0.2 ml per served in the 11 patients presenting with end-stage chronic renal well) in RPMI 1640 medium containing antibiotics, and 10% failure before the institution of dialysis replacement therapy FCS and 1 or 3 U/mI of recombinant IL-2 (Sandoz, Wien, was significantly decreased as compared to normal controls' Austria). Proliferation was assayed after 48 hours of culture by values (mean SEM 0.85 0.24 vs. 2.49 0.22 U/mI; Fig. 2). means of a 16-hour [3H] thymidine incorporation assay (CEA, Six out of these 11 patients (50%) presented an IL-2 activity 0.3 U/mI, that is, levels close to background. Gif-sur-Yvette, France).

Beaurain et a!: In vivo T cell activation and uremia

• •

0.9

• . 0.6. 0.4

•

r = 0843

•

:

•

0.2

p<

..

0 0

0.2

0.4

0.6

0.8

1

1.2

• •

22.4, P < 0.001. The difference was not significant when comparing the indexes observed in non-hemodialyzed and

S

•

S

V

known to possess variable affinity for the IL-2 molecule, namely, high and low affinity IL-2 receptors [36]. Firstly, immunofluorescence labeling using the IOTl4a monoclonal

• •

2

• • •

1

0.85

• •

A

S. 0.24

• : 50.93

-U•••

0.12

S

antibody was used to detect the Tac antigen expressed by the a chain of the IL-2 receptor that is shared by both high and low affinity receptors [36]. Double staining using a mixture of OKT3 (that specifically recognizes T cells) and IOTl4a allowed the direct detection of T cells bearing the Tac antigen. Secondly, since only high affinity receptors trigger cell proliferation following IL-2 binding, such receptors were studied by evaluating

the capacity of non-stimulated E T cells to proliferate in

••• • •• B

exclude that any soluble factor present in the analyzed supernatants interfering with the CTLL-2 biological assay was responsible for the low IL-2 activities detected. Further stressing this point the addition of serial dilutions of culture supernatants from uremic patients presenting low IL-2 activities, to recombinant or natural IL-2 reference preparations did not modify their IL-2 activity as assessed by the CTLL-2 assay. Effect of dialysis session. In nine of the uremic hemodialyzed patients studied, PHA-induced IL-2 activity and cell proliferation were assessed in cultures from T cells collected before and just after a dialysis session. No significant difference was noted when comparing the data by means of the Student's t-test for paired data (Table 1).

Freshly collected E T cells that were not submitted to any prior in vitro stimulation were studied for their IL-2 receptor expression. Two different methods were used in parallel allowing the detection of two distinct types of IL-2 binding sites

0.22

1

the biological assay or the soluble phase RIA allowed to

Spontaneous expression of IL-2 receptor by peripheral T cell (in the absence of mitogen activation)

3

0

alysis was similar to the one found in non-hemodialyzed patients (mean SEM 0.93 0.12 vs. 0.85 0.24 U/mI; Fig. 2) with, here, again a significant decrease compared to normal controls' values (2.49 0.22 U/mI; Fig. 2). When compared to normal controls, uremic hemodialyzed patients showed a significantly decreased mean PHA-induced cell proliferation index: mean SEM 79.5 17.2 versus 184.8 hemodialyzed patients. The fact that similar results were observed when using either

Biological assay, U/mi Fig. 1. Correlation between values observed for IL-2 levels in culture supernatants from uremic patient's stimulated T cells as assessed by a biological assay using the CTLL-2 line and a radioimmunoassay using human iodinated IL-2. Results were expressed in U/mI as referred to 2 different reference preparations (Methods).

4.

639

presence of exogenous recombinant IL-2. The proportion of normal freshly collected E cells expressC

Fig. 2. Interleukin-2 activity in stimulated T cell culture supernatants.

ing the Tac antigen was 2.15 0.38% (mean SEM; Fig. 3). After short term (48 hrs) in vitro exposure of cells to exogenous

recombinant IL-2 Tac antigen was expressed on 3.17 0.80% of the cells recovered (Fig. 3). Results obtained when evaluattients (column C). For each group of patients the horizontal bars ing thymidine incorporation under the same culture conditions represent mean values. The detailed mean SEM values are reported are detailed in Figure 4. for all the 3 groups analyzed. Uremic non-hemodialyzed patients. Freshly collected E rosetting lymphocytes from uremic non-hemodialyzed patients showed a proportion of Tack T cells that was significantly In these patients mean PHA-induced cell proliferation index increased as compared to normal controls' values mean SEM: Values observed in normal controls (column A); uremic non-hemodialyzed (non-HD) patients (column B); uremic hemodialyzed (HD) pa-

was decreased although not significantly as compared to normal controls: 116.1 39.0 versus 184.8 22.4 (mean sEM). Hemodialyzed patients. The IL-2 activity detected in culture supernatants from uremic patients undergoing regular hemodi-

5.69 0.63% versus 2.15 0.38% (Fig. 3). Moreover, the spontaneous proliferative response of these T

cells was also significantly higher than the one recorded in normal controls (Fig. 4). Thus, 6 out of the 20 patients studied

640

Beaurain el a!: In vivo T cell activation and uremia

Table 1. Effect of the hemodialysis session on peripheral T cell phenotypes a nd function % CD4+/%CD8+ ratio IL-2 activity Mitogenic response Pre-HD Post-HD Pre-l-ID Post-HD Post-HD Pre-HD

Patients CHA SAL IPP ROB MON

NDA BAY BAU KOU

1.4

1.3

191.4

70.8

1

1

227.5

251.9

0.5

0.7

278.1

164.2

2.4 0.6 0.6

2.1

63.0

0.5 0.6

58.1 24.3

1.4 0 1.3

32.7 131.4

1.1

0 1.5

17.3

0.81 0.81 1.37

0.86

1.05

47.1 97.2 32.8

2.54

2.61

33.8

2.06

66.9 25.8

1.02 1.35

1.61

—

—

1.23

1.46 1.8

0.94 1.25

For all parameters no significant difference between pre-HD and post-HD values was assessed by using the Student's t test for paired data.

•t49.4 30 -

. S

.

A

, 20

...

a

I0

.

Co

I10 — A

A

.1 ti. ..

r

•

ta At

£ £ a A

a£ £

•

4. A

AA

0—

A assasa ...n.

. — Fig. 3. Percentage values of T cells

I

*

La

At

At a

I

I

I

I_

I

I

I____________ I I

a

b

C

a

b

c

a

Normal controls

Non-HO uremic patients

b

c

HO uremic patients

showed a thymidine incorporation at 48 hours of culture in medium alone (in the absence of any stimulation) that exceeded the mean value 2 SD exhibited by normal T cells. When these

expressing the IL-2 receptor a chain as assessed by labeling with an anti-Tac monoclona! antibody. Results are detailed for normal controls, uremic non-hemodialyzed (non-HD) and hemodialyzed (HD) patients. Cells were studied as collected from peripheral blood at time 0 (column a) or following 48 hrs of culture in presence of 0 U/mi (column b) or 2 U/mI (column C) of human recombinant IL-2. The horizontal bars represent mean values that are as follows: Normal controls: (a) 2.15 0.38%; (b) l.86 0,80%. Non-HD uremic 0.41%; (c) 3.17 0.38%; patients: (a) 5.69 0.63%; (b) 5.35 (c) l2.14 0.95%. HO uremic patients: (a) 7.11 0.81%; (b) 9.58 1.41%; (C) 19.08 1.98%.

revealed an abnormally high sensitivity of non-stimulated hemodialyzed patients' T cells to the lymphokine, with a mean

proliferation that was significantly higher in hemodialyzed patients' cells were cultured for 48 hours in presence of patients as compared to both normal controls and uremic recombinant IL-2, a significantly higher response than the one non-hemodialyzed patients (Fig. 4). At variance with nonfound with normal E cells was noted (Fig. 4). Increasing the hemodialyzed patients a dose-dependent IL-2 induced proliferamounts of exogenous IL-2 added to the cultures did not ative response was observed in hemodialyzed patients (Fig. 4). Effect of the hemodialysis session. As detailed on Table 2 the significantly influence the proliferative response (Fig. 4). Uremic hemodialyzed patients. The proportion of peripheral dialysis session by itself did not influence the expression of IL-2 Tact T cells in uremic hemodialyzed patients was significantly receptor on hemodialyzed patients' peripheral T cells. Thus, similar phenotypic and functional patterns were found when increased as compared to normal controls (mean SEM 7.11 0.81% vs. 2.15 0.38%). The spontaneous proliferative re- analyzing peripheral T cells in nine different patients before and sponse of hemodialyzed T cells was also significantly higher after the hemodialysis session. None of the parameters menthan the one observed in normal controls (Fig. 4). This mean tioned above namely, Tac expression either on fresh T cells or

spontaneous proliferation was also higher than the one observed with uremic non-hemodialyzed patients' E cells, al-

on T cells cultured in presence of IL-2, spontaneous T cell proliferation and the increased sensitivity to exogenous IL-2

though in this case the difference was not statistically significant

showed any significant modification when analyzed by means of the Student's I-test for paired data (Table 2).

(Fig. 4). Lastly, addition of exogenous IL-2 to the cultures

641

Beaurain et a!: In vivo T cell activation and uremia

Analysis of peripheral T cell subset phenotypes 14

S

The percentage values of total T cells (CD3), helper-inducer (CD4) and cytotoxic/suppressor (CD8) T cells was analyzed by labeling of E rosetting T lymphocytes with specific monoclonal antibodies (that is, OKT3, OKT4 and OKT8). The % CD4/% CD8 ratio was 1.37 0.11 (mean SEM) in normal controls, 2.10 0.58 in uremic non-hemodialyzed patients and 1.51 0.12 in uremic hemodialyzed patients. Three values without any statistically significant difference when compared by using the Student's t-test. Furthermore, as detailed in Table I, the dialysis session did not provoke, at least in the nine patients analyzed, any signif-

•

icant acute modification in the proportions of the different T cell

12 10 -

•

8

•

0

•

x 6

+

6

•

C

••

• •+ S

4

4+

••

•

i-f

2

• S

•

:

3

0

•

:

1

3

Normal controls

0

1

Non-ND uremic patients

pared to controls (data not shown) but, when results were expressed as percentage values the proportions between the main regulatory T cell subsets were normal.

0 0

subsets analyzed. Uremic patients included in this study were lymphopenic, a finding in complete accordance with several other reports [1—3, 38], including ours [27]. Thus, absolute counts of the various T cell subsets were decreased as com-

1

3 IL-2(U/ml)

HD uremic patients

4. Pro!ferative capacity of peripheral blood T cells from normal controls, uremic non-hemodialyzed (non-HD) and hemodialyzed (ND) Fig.

patients cultured for 48 hrs in presence of I or 3 U/mI of human recombinant IL-2. Results are expressed as counts per minute (cpm). The horizontal bars represent mean values that are as follows (mean SEM):

Normal controls o U/mI 207.6 22.6 (1) I U/mI 1647.5 244.5 (2) 3 U/mI 2508.3 346.3 (3)

Non-RD uremic patients 0 U/mI 402.6 59.6 (4) I U/mI 3501.5 349.4 (5) 3 U/mI 3915.3 386.9 (6) IID-uremic patients 0 U/mI 708.3 151.4 (7) I U/mI 5438.0 730.0 (8) 3 U/mI 7075.3 810.1 (9)

Statistical comparison: 1 and 4) P < 0.003; I) and 7: P < 0.003; 2) and 5): P < 0.0002; 2) and 8): P < 0.0001; 3) and 6): P < 0.01; 3) and 9): P

<0.003; 6) and 9): P < 0.003.

Soluble circulating IL-2 receptor We detected a soluble form of the Tac antigen in the patient and control plasma by means of a specific ELISA. The physicochemical properties of this soluble form of the IL-2 receptor, shown to be released from different types of normal or tumoral

Discussion

Phenotypic and functional parameters related to T cell activation have been analyzed in a series of clinically homogeneous

uremic patients undergoing or not chronic hemodialysis. We concentrated our attention on the production and receptor targeting of a major T cell derived lymphokine, namely interIeukin-2 (IL-2), since our preliminary reported data [27] had indicated that both these parameters were severely affected in hemodialyzed patients. Moreover, IL-2 is essential for appropriate cell cooperation in delayed type hypersensitivity reactions known to be greatly impaired in uremic patients. Thus, any abnormality in the bioavailability of IL-2 may represent a major pathophysiological element to explain the secondary immunodeficiency associated with chronic uremia. Results obtained in this study totally confirm on a large series of 23 hemodialyzed patients the observations made in our first report [27] showing abnormally increased proportions of circulating T cells spontaneously expressing the IL-2 Rec. IL-2 Rec is expressed at the surface of activated T cells under different molecular forms. Binding studies using radiolabeled IL-2 have demonstrated that T cells may express at least two sorts of IL-2 binding sites presenting different affinities for IL-2, that is, high (Kd = 10 M) and low (Kd = 10—8 M) affinity

receptors [36]; only high affinity receptors can trigger IL-2 mediated cell proliferation [36, 39]. Biochemical studies allowed the characterization of two different polypeptide chains termed a (55 kDa) and /3 (75 kDa) [40—431. Low affinity receptors correspond to the exclusive expression of the a chain

Tack cells have already been reported [34, 37]. Results are at the cell surface; IL-2 binding to these sites do not lead to expressed in U/mi according to a reference curve set up by endocytosis of the growth factor-receptor complex [43]. At using as a standard a cell free supernatant of an IL-2 dependent variance, both a and /3 chains (noncovalently linked) are needed human T cell line. for the expression of high affinity receptors able to mediate IL-2 As shown on Figure 5 uremic non-hemodialyzed and hemo- internalization [39—43]. Exclusive expression of the /3 chain (a dialyzed patients exhibited circulating levels of soluble IL-2 situation only described for some cell lines [43]) can mediate receptor that were significantly increased as compared to nor- internalization of IL-2 and cell growth [43]. Only the a chain mal control values. In six cases, plasma samples were analyzed expresses the Tac epitope identified by the anti-Tac or similar before and after the dialysis session and no major modification monoclonal antibodies [29]. Thus these monoclonal antibodies of serum soluble IL-2 receptor was noted (Fig. 5). will target both high and low affinity receptors. Thus, by

642

Beaurain et a!: In vivo T cell activation and uremia

Table 2. Effect of the hemodial ysis session on preactivated peripheral T lymphocytes Sensitivity to exogenous IL-2

Tac T cells % circulating Tack cells Pre-HD Post-HD

Patients CHA SAL IPP ROB MON NDA BAY

BAU KOU

13.8

14.6

13.5

12.1

7.8 16.7 6.6 15.2 7.3 7.7 7.6

12.8 19.5

Proliferation cpm

%

Pre-HD

Post-HD

Pre-HD

Post-HD

34.1 18.9

26.6

9132 8962 3242 6278

8220 5334 4728 7016 9500

23.4 49.3 22.9 21.0

23 29.3 54

7.4 12.8

6.3 8.6

15.6

25.6 20.7 14.4 24.7 33.3

27.0 28.2

12.4

9651 12485

15104 10038

6108

13934 14682 12007 7953

For all parameters no significant difference between pre.HD and post-HD values was assessed by using the Student's t-test for paired data.

recombinant IL-2 alone, one may infer which type of IL-2 receptor (high or low affinity) is expressed by a given population. Direct binding studies using radiolabeled IL-2 were impossible to perform on hemodialyzed patients due to the high

...'..

20000 —

A

number of purified IL-2 Rec cells that would have been needed. The results obtained confirmed that in hemodialyzed patients peripheral T cells that spontaneously express the Tac antigen also exhibit an abnormally increased sensitivity to exogenous

10000 —

5000 —

A

IL-2. Thus, at variance with what was observed in normal

At

controls, peripheral T cells from hemodialyzed patients are able to proliferate in short term cultures in presence of recombinant IL-2. This increased proliferation is fully paralleled, in the same

A

I

cultures, by the presence of impressively high proportions of IL-2 Rec positive cells. These results strongly suggest that, at

least part of this particular T cell subset is spontaneously

A

expressing in vivo high affinity IL-2 Rec.

1000 —

The plasma of the same patients was analyzed for the presence of the soluble form of the Tac antigen since an

.

increased IL-2 Rec expression may lead to its release in the circulation. In all hemodialyzed patients studied, levels of

500 —

soluble Tac antigen were increased 10- to 20-fold as compared 300 —

100 —

to normal controls. This increase in circulating Tac may be related to the increased expression of Tac by T cells, although

+ Normal controls

one cannot exclude that soluble Tac also represents a retention product. Although this soluble form of the IL-2 Rec is able to bind IL-2 [44], it is not known whether it represents a marker of Non-HO uremic patients

Before

After

HD

HD

abnormal T cell activation or if it can really interfere with

are reported whose serum was tested before and after the dialysis

regulatory processes. A significantly decreased IL-2 production was found in 48 hours culture supernatants from stimulated hemodialyzed patients' T cells; 56% of patients showed IL-2 levels < lU/mi. Similar data, that are concordant with our previous results [27], have also been described by other groups [17, 18, 45]. Since identical results were obtained by using either the conventional

session. The horizontal bars represent mean values that are as follows: Normal controls, 253.5 29.3 U/mI; Non-lID uremic patients, 3220.7 832.4 U/mI; HD uremic patients, 4036.0 947.1 U/mI.

biological IL-2 assay (using the murine, CTLL-2 line) or a soluble phase RIA (using radiolabeled human IL-2) one may

HD uremic patients

FIg. 5. Circulating soluble IL-2 receptor levels as detected by means of a specific ELISA in the serum of normal controls, non-hemodialyzed (non-HD) or hemodialyzed (HD) patients. Six hemodialyzed patients

exclude the possibility that the defect observed could be related to a soluble inhibiting material present in the culture supernatants and interfering with the biological assay. Moreover, we

applying in parallel an immunofluorescence assay with an have also shown that the significantly lower IL-2 activity anti-Tac monoclonal antibody and a functional test defining the observed is not related to an excessive production of proscapacity of Tack cells to proliferate in presence of exogenous taglandins E2 (PGE2) by remnant monocytes present in the

643

Beaurain et a!: in vivo T cell activation and uremia

culture [27] (monocytes can depress IL-2 production by releas-

Acknowledgment

ing PGE2 which in turn activate suppressor I lymphocytes The radioimmunoassay kit used in this study was a gift from [46]). This result was confirmed by others using a different Medgenix, Fleurus, Belgium. technical approach [451. The abnormal mitogen-induced 11-2 activity could be related either to defective production and/or to increased absorption of the lymphokine by activated Tack T cells. Supporting this latter hypothesis are the results showing that, even in low amounts, IL-2 significantly up-modulates IL-2 Rec expression in hemo-

dialyzed patients' T cells, establishing a vicious circle. However, if absorption was the only mechanism one would expect

to observe either a normal or an increased PHA induced proliferation. Since this is not the case namely, a decreased response is generally observed, the presence of an associated T cell defect cannot at this point be ruled out.

Recently it has been suggested that the impaired T cell response could be related to monocyte dysfunction [181. However, results concerning this issue are still controversial. Some

authors describe normal or increased IL-I production from stimulating monocytes in hemodialyzed patients [47, 48] while others found defective production [49]. Moreover, some data may also suggest that monocyte related factors, different from IL-l, could also be implicated [18]. Results obtained when extending the study to a series of 20

Reprint requests to Lucienne Chatenoud, INSERM U 25, Hôpital

Necker, 161 rue de Sèvres, 75743 Paris Cedex 15, France,

References I. REVILLARD JP: Immunological alterations in chronic renal insufficiency. Adv Nephrol 8:365—382, 1979

2. MILLER TE, STEWART E: Host immune status in uremia. I. Cell mediated immune mechanisms. Clin Exp immunol 41:115—122, 1980 3. NAKHLA LS, GOGGIN MJ: Lymphocyte transformation in chronic renal failure. Immunology 24:229—235, 1973 4. KIRKPATRICK CH, WILSON WEC, TALMAGE DW: Immunologic studies in human organ transplantation. I. Observation and characterization of suppressed cutaneous reactivity in uremia. J Exp Med 119:727—742, 1964

5, DAMMIN GJ,

COUCH NP, MURRAY JE: Prolonged survival of skin homografts in uremic patients. Ann NY Acad Sci 64:967—976, 1957 6. MoRRIsoN AB, MANESS K, TAWES R: Skin homograft survival in chronic renal insufficiency. Arch Pathol 75:139-143, 1963

7. CROSNIER J, JUNGERS P. COUROUCE AM, LAPLANCHE A, BENHAMOU E, DEG0S F, LACOUR B, PRUNET P. CERISIER Y, GUESRY

P: Randomised placebo-controlled trial of hepatitis B swface

antigen vaccine in French haemodialysis units: II. Haemodialysis patients. Lancet i:797—800, 1981 8. KOHLER H, ARNOLD W, RENSCHIN 0, DORMEYER HH, MEYERZUM BOSCHENFELDE KH: Active hepatitis B vaccination of dialysis patients and medical staff. Kidney mt 25:124—148, 1984

patients presenting end stage uremia before the beginning of hemodialysis revealed that similar features of T cell activation were present, although to a lesser degree than in hemodialyzed 9. STEVENS CE, ALTER MD, TAYLOR PE, ZANG EA, HARLEY EJ, patients. Thus, peripheral Tact T cells and circulating IL-2 Rec SZTIUNESS W: Hepatitis B vaccine in patients receiving hemodiallevels were abnormally high in these patients as compared to ysis. Immunogenecity and efficacy. N EngI J Med 311:496—500, 1984 normal controls, although the values were lower than those W, STEVENS CE, HARLEY EJ, ZANG EA, OLESZKO WR, observed in hemodialyzed patients. Similarly, increased T cell 10. SzMUNE55 WILLIAM DA, SADOVSKY R, MORRISON JM, KELLER A: Hepatitis sensitivity to exogenous IL-2 was evidenced that was, howB vaccine: Demonstration of efficacy in a controlled clinical trial in ever, less striking than the one exhibited by uremic hemodiaa high risk population in the United States. N EngI I Med 303: 833—841, 1980 lyzed patients. Finally, in uremic non-hemodialyzed patients R, VAN BEERS D, LIESNARD C, DRATwA M: Impaired IL-2 activity in stimulated T cell culture supernatants and in 11. CAPPEL humoral and cell-mediated immune responses in dialyzed patients association with a subnormal PHA proliferative responsiveness after influenza vaccination. Nephron 33:21—25, 1983 was decreased. 12. ANDREW OT, SCHOENFELD PY, HOPEWELL PC, HUNPHREYS MH: Tuberculosis in patients with end-stage renal disease. Am I Med These results suggest that even before starting renal replace68:59—65, 1980 ment therapy by means of hemodialysis, the immunodeficiency 13. LUNDIN AP, ADLER AJ, BERLYNE GM, FRIEDMAN EA: Tubercuof chronic uremia is already associated with phenotypical and losis in patients undergoing maintenance hemodialysis. Am J Med functional signs of T cell preactivation. Moreover T cell preac67:597—602, 1979 tivation is significantly accentuated by hemodialysis, indepen- 14. GOLDBLUM E, REED WP: Host defenses and immunologic alter. ations associated with chronic hemodialysis. Ann intern Med dently from the kind of membrane used. 1980 No major modifications could be evidenced in any of the 15. 93:597—613, SINGH S, HURTUBISE P, MICHAEL 0, PESCE A, POLLAK V: phenotypical or the functional tests performed in the nine Preliminary observations on the laboratory markers of cell-mediuremic hemodialyzed patients were studied prior to the onset ated immunity in patients transferring from hemodialysis to continand at the end of the hemodialysis session. This is at variance uous ambulatory peritoneal dialysis, in Immunological Changes in Hemodialysis. Basel, Karger, Contr Nephrol 36:73—81, 1983 with other reports in the literature describing modifications in E, LADEFOGED J: Cellular immunity in renal failure: the proportions of the various T cell subsets [50], or of their 16. LANGHOFF Depression of lymphocyte transformation by uraemia and methylfunctional capacity [51] during or immediately after the dialysis

session. It is difficult to give any reliable explanation for this

discrepancy although one may assume that variabilities in patient selection represent the major factor. To conclude, the data presented in this work suggest that uremia per se induces a state of in vivo T cell preactivation whose intensity is exacerbated by hemodialysis. Moreover, they underline that the abnormality of IL-2 production and receptor targeting, observed both in uremic hemodialyzed and non-hemodialyzed patients, is central to the state of immunodeficiency induced by chronic renal failure.

prednisolone. Intra-individual consistency of lymphocyte responses to the in vitro suppressive effect of steroid. mt Arch Allergy AppI immunol 74:241—245, 1984 J, GHOBRIAL I, SHEA SM, EBERT EC, EI5INGER RP,

17. RASKOVA

K JR: T cells in patients undergoing chronic hemodialysis: Mitogenic response, suppressor activity and interleukin 2 production and receptor generation. Diagnostic immunol 4:209—216, 1986 RASKA

18. MEUER SC, HAUER M, Kuitz P, MEYER ZUM BUSCHENFELDE KH,

KOHLER H: Selective blockade of the antigen receptor mediated

pathway of T cell activation in patients with impaired primary immune responses. J Clin Invest 80:743—749, 1987 19. KURZ P, KOHLER H, MEUER SC, HUTrEROTH TH, MEYER ZUM BUSCHENFELDE KH: Impaired cellular immune responses in

644

Beaurain ci a!: In vivo T cell activation and uremia

chronic renal failure: Evidence for a T cell defect. Kidney ml

released from activated human lymphoid cells in vitro. J

29:1209—1214, 1986

135:3172—3177, 1985

20. LORTAN JE, KIEPIELA P, COOVADIA HM, SEEDAT YK: Suppressor

cells assayed by numerical and functional tests in chronic renal failure. Kidney In! 22:192—197, 1982 21. TSAKOLOS ND, THEOHARIDES TC, HENDLER ED, GOFFINET J, DWYER JM, WHISLER RL, ASKENASE PW: Immune defects in

chronic renal impairment: Evidence for defective regulation of

RUBIN LA, KURMAN CC, FRITZ ME, BIDDISON WE, GOLDMAN ND, NELSON D: A monoclonal antibody, 7G7/B6, that binds to an epitope on the human IL-2 receptor distinct from that recognized by IL-2 or anti-Tac. Hybridoma 4:91—102, 1985 36. ROBBRi, GREENE WC, RUSK CM: Low and high affinity cellular receptors for interleukin 2. Implications for the level of Tac antigen. 35.

JExp Med 160:1126—1146, 1984

lymphocyte response by macrophages from patients with chronic renal impairment on haemodialysis. C/in Exp Immunol 63:218—227,

37. Rusir LA, KURMAN CC, FRITZ ME, YARCHOAN R, NELSON DL: Identification and characterization of a released form of the IL-2

1986

in Leucocytes and Host Defense, edited by JJ OPPENHElM, DM JACOBS, New York, Alan R. Liss Inc, 1985, p. 95 38. RASKA K, RASKOVA J, SHEA SM, FRANKEL RM, WOOD RH, receptor,

22. ALEVY YG, MUELLER KR, ANDERSON JR. HUTCHESON PS, SLAVIN RG: The role of monocytes and responder cells in suppression of antigen-specific I cell proliferation in chronic renal failure. mt Arch Allergy AppI Immunol 73:97—103, 1984 23. KLEIMAN KS, ZOSCHKE DC: Suppressionof human lymphocyte responses in chronic renal failure mediated by adherent cells: Analysis of serum-free media. J Lab C/in Med 106:262—268, 1985 24. RASKOVA J, GHOBRIAL J, SHEA SM, EISINGER RP, RASKA K JR: Suppressor cells in end stage renal disease. Functional assays and monoclonal antibody analysis. Am J Med 76:847—853, 1984

LIFTER J, GHOBRIAL I, EI5INGER RP, HOMER L: T cell subsets and

cellular immunity in end-stage renal disease. Am J Med 75:734— 740,1983

39. ROBB Ri, GREENE WC: Internalization of interleukin 2 is mediated by the /3 chain of the high-affinity interleukin 2 receptor. J Exp Med 165:1201—1206, 1987

40. SHARON M, KLAUSNER RS, CULLER BR, CHIzz0NITE R, LEON-

ARD Wi: Novel interleukin 2 receptor submit detected by cross-

Humoral inhibitors of the immune response in uremia. V. induction of suppressor cells in vitro by uremic serum. AmfPaihol 111:149—155, 1983

linking under high affinity conditions. Science 234:859—863, 1986 41. TSUDO M, KOZAK RW, GOLDMAN CK, WALDMAN TA: Demon-

RASKA K JR, Morrison AB, Raskova J: Humoral inhibitors of the immune response in uremia. III. The immunosuppressive factor of uremic rat serum in a very low density lipoprotein. Lab Invest

Nat! Acad Sci USA 83:9464—9698, 1986 42. KUKOVICH M, WANO Y, LE Tm BICH THUY, KATZ P. CULLEN BR, KEHRL JH, GREENE WC: A second human IL-2 binding

25. RASKOVA J, RASKA

26.

Immunol

K

JR:

42:636—642, 1980

27. CHATENOUD L, DUGAS B, BEAURAIN G, TOUAM M, DRUEKE 1, VASQUEZ A, GALANAUD P, BACH JF, DELFRAISSY JF: Presence of preactivated T cells in hemodialyzed patients: Their possible role in altered immunity. Proc Nat! Acad Sci USA 83:7457—7461, 1986 28. WATSON J, MocHizuKi D: Interleukin 2: A class of T cell growth factors Immunol Rev 5 1:257—278, 1980 29. UCHIYAMA T, BRODER 5, WALDMAN TA: A monoclonal antibody

(anti-Tac) reactive with activated and functionally mature human T cells. I. Production of anti-Tac monoclonal antibody and distribution of Tac (+) cells. J Immunol 126:1393—1397, 1981 30. REINHERZ EL, SCHLOSSMAN SF: The differentiation and function of human T lymphocytes. Cell 19:821—827, 1980

31. CHATENOUD L, BACH MA: Abnormalities of T cell subsets in glomerulonephritis and systemic lupus erythematosus. Kidney In! 20:267—274, 1981

32. M05sMAN T: Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxic assays. J immunol Meih 65:55—63, 1983 33. BERTOGLIO J, floissor N, BONNET MC, CHATENOUD L, CHOUAIB 5, DEGIOVANNI G, FRADELIZI D, GODARD A, GOUBE DE LA FORET P, HAREL-BELLAN A, JAcuBovucH R, LACAZE M, LAHAYE T, M0u5ALAYI M, PLA M, REY A, ROTHSCHILD E, ScHMirr C,

stration of a non-Tac peptide that binds interleukin 2: A potential participant in a multichain interleukin 2 receptor complex. Proc

protein

that may be a component of high-affinity IL-2 receptors.

Nature 327:518—522, 1987

43. Tsiicw K, WARF HM, KATO K, SMITH KA: Interleukin 2 high-affinity receptor expression requires two distinct binding proteins. J Exp Med 165:223—238, 1987 44. RUBIN LA, JAY G, NELSON DL: The released interleukin 2 receptor binds interleukin 2 efficiently. J Immunol 137:3841—3844, 1986 45, LANGHOFF E, HOFMANN B, ODUM N, LADEFOGED J, PLATZ P. RYDER LP, SVEJGAARD A: Kinetic analysis of interleukin 2 (IL-2) production and expression of iL-2 receptors by uraemic and normal lymphocytes. Sand J Immunol 25:29—36, 1987 46. CHOUAIB S. FRADELIZI D: The mechanism of inhibition of human IL-2 production. J Immunol 129:2463—2468, 1982 47. LUGER A, KOVARIK J, STUMMVOLL HK, URBANSKA A, LUGER

TA: Blood-membrane interaction in hemodialysis leads to increased cytokine production. Kidney In! 32:84—88, 1987 48. SHALDON S, DINARELLO CA, KOCH KM. LONNEMANN 0, BINGEL M, GRANOLLERAS C, DESCHODT G, BRANGER B, OULES R, FouRCADE J: Interleukine 1 et dialyse, in Actualités Nephrologiques de

l'Hôpiial Necker, edited by JP GRUNFELD, J CROSNIER JL FUNCKBRENTANO, iF BACH Paris, Flammarion, 1987, pp. 391—402 49. BLUMENSTEIN M, SCHMIDT B, WARD RA, ZIEGLER-HEITBROCK

HWL, GURLAND Hi:

Altered

interleukin 1 in patients undergoing

Workshop on Standardization of human iL-2: Joint report. Lym-

hemodialysis. Nephron 55:277—281, 1988 CHIDA Y, SAKURAI 5, YOSHIYAM N: The effect of hemodialysis on lymphocyte subsets during dialysis. Gun Nephrol25:159—163, 1986

phokineRes 1:121—127, 1982 34, RUBIN LA, KURMAN CC, FRITZ ME, BIODI50N WE, BOUTIN B, YARCHOAN R, NELSON DL: Soluble interleukin 2 receptors are

51. DE GAST GC, PONDS E, ARITZ L, THE TH, VAN DER HEM 0K: Impaired lymphocyte function and neutrophil damage during the first hours of hemodialysis. C/in Nephrol 8:514—519, 1977

SERROU B, SouLILLou JP, TRIBEL F, YTHIER A: First French

50.