European Journal of Clinical Investigation ( 1990) 20, 555-562

Chemotaxis of the peritoneal cells and the detection of a chemo-attractant in the effluent from peritoneal dialysis patients H. J. BOS*, D. M. BOORSMA?, C. W. TUK*, J. C. DE VELD*, A. J. C. VAN DER MUYSENBERG*, T. J. M. HELMERHORSTS, D. G. STRUIJK**, H . VAN BRONSWIJKg & R. H. J . BEELENT, Departments of *Cell Biology, t Dermatology $Obstetrics and Gynaecology, §Internal Medicine, Division of Nephrology and 1' Haematology, Medical Faculty and Universal Hospital, Free University, Amsterdam, and ** Department of Nephrology, University Hospital, Academic Medical Centre, Amsterdam, The Netherlands Received 12 November 1988 and in revised form 6 December 1989

Abstract. The migration of peritoneal cells from 25 continuous ambulatory peritoneal dialysis patients and eight healthy women undergoing laparoscopy were studied. Peritoneal cells of continuous ambulatory peritoneal dialysis patients migrated to commonly used chemoattractants, like N-formyl-methionyl-leucyl-phenyl-alanine-methyl-esterand casein, but they also migrated to high concentrations of recombinant interleukin- I CI and to endotoxin (lipopolysaccharide). In the peritoneal effluent from continuous ambulatory peritoneal dialysis patients a rather heat stable chemoattractant was found with a molecular weight of 40200 kDa with an optimal activity at approximately 125 kDa. The chemoattractant is a protein and is not complement factor 5a or interleukin-I and was only found in peritoneal effluent from continuous ambulatory peritoneal dialysis patients, but not in peritoneal fluid from healthy women undergoing laparoscopy. Therefore, peritoneal dialysis might induce the generation of a chemoattractant. The optimal chemotactic response of peritoneal cells from continuous ambulatory peritoneal dialysis patients to N-formylmethionyl-leucyl-phenyl-alanine-methyl-ester in medium could be enhanced by replacing the medium by peritoneal effluent. So the chemotaxis of peritoneal cells to the factor in the peritoneal effluent is caused by another mechanism, which might involve different cell Abbreviations: AL. aldolase; BSA. bovine serum albumin; C5a. complement factor 5a; CAPD. continuous ambulatory peritoneal dialysis; CII. chymotrypsin; F, ferritin; FCS. foetal calf serum; FMLP. N-formyl-methionyl-leucyl-phenyl-alanine-methyl-ester~ HPLC, high-pressure liquid chromatography; IL-I, interleukin-I; LAF. lymphocyte activation factor; LPS. lipopolysaccharide; PC, peritoneal cells; PE. peritoneal effluent; PVP-F. polyvinylpyrrolidone-free; rlL-lx, recombinant interleukin-lz; uRPM1 1640 medium with l0'%,heat inactivated FCS-60 min. 56 C (Gibco BioCult, Irvine, UK). 2 mmol I - ' L-glutamine. 50 u nil-' penicillin, 50 j i g ml streptomycin (Mycoform, Delft. The Netherlands).

'

Correspondence: H . J. Bos. Department of Cell Biology, Division of Electronmicroscopy. Medical Faculty. Free University. PO Box 7161. NL-1007 MC Amsterdam. The Netherlands.

surface receptor populations, than the chemotactic response to N-formyl-methionyl-leucyl-phenylalanine-methyl-ester.

Keywords. Chemoattractant, chemotaxis, interleukin I , lipopolysaccharide, macrophages, peritoneal dialysis, peritoneal effluent. Introduction Continuous peritoneal dialysis is an alternate treatment for haemodialysis of chronic uraemic patients. However, bacterial peritonitis is still a serious complication that occurs in relatively high numbers of these patients and frequent infections cause significant morbidity [ 1,2]. The local defence mechanism of peritoneal dialysis patients might play a key role in the prevention of bacterial peritonitis and it is generally accepted that phagocytic cells determine the local defence. Therefore, the migration into the peritoneal cavity and the phagocytosis and killing of microorganisms by peritoneal cells (PC) is essential for the prevention of a bacterial peritonitis. The mononuclear phagocyte is the predominant cell-type found in peritoneal dialysate from uninfected patients undergoing peritoneal dialysis [3-51as well as in peritoneal fluid from normal and infertile women [6-91. These findings support an important role for macrophages in the local defence mechanism and so for the prevention of a bacterial peritonitis in continuous ambulatory peritoneal dialysis (CAPD) patients. Therefore, we investigated the in ziitro migration of PC from uninfected CAPD patients and healthy women undergoing laparoscopy. In addition we determined and partially characterized a chemoattractant in the peritoneal effluent (PE) from CAPD patients. The significance of the chemotaxis of PC from CAPD patients and the origin and relevance of the chemoattractant present in the effluent for the prevention of a peritoneal infection is discussed.

555

556

H. J. BOS et ul.

Patients and methods

Source qf’tiie peritoneul cells

We studied PC from 25 CAPD patients, 15 male and 10 female, ranging from 24 to 74 years of age. All patients were studied when uninfected and no patient was studied within 1 month of an episode of peritonitis. PC were obtained by centrifugation (300g, 10 min, 4’C) of the complete dialysate effluent volume ( 1 to 2.5 1) after an overnight dwell time of at least 8 h. Eight healthy volunteers were selected from among women undergoing elective laparoscopy for diagnostic evaluation of infertility or tuba1 sterilization. Patients with endometriosis, pelvic inflammatory disease or neoplasia were excluded because of reported macrophage abnormalities [9]. PC were obtained immediately following placement of the laparoscope and prior to any surgical manipulation. Peritoneal fluid was aspirated by gentle manual suction under direct visual control from the cul-de-sac. The cell-rich fluid (10-1 5 ml) was processed in the same fashion as the peritoneal effluents from the CAPD patients. Sterile conditions were maintained throughout these procedures. The obtained cells were counted in Biirker chambers and viability was determined by trypan blue exclusion. Differential cell counts were determined from cytocentrifuge preparations stained with May-GrunwaldGiemsa and a minimum of 200 cells were counted. PC were washed three times in aRPMI [RPMI 1640 medium with 10% heat-inactivated foetal calf serum (FCS), 60 min, 56 C (Gibco Bio-Cult, Irvine, UK) 2 mmol I - ’ L-glutamine, 50 u ml-’ penicillin, 50 pg ml-’ streptomycin (Mycoform, Delft, The Netherlands)] before testing in the chemotaxis assay. Diulysis ,fluid. peritoneul eguent, concentrution und Jructionution of the peritoneul cjluents

Fresh commercial dialysis fluids (Travenol Laboratories Ltd, Thetford, Norfolk, UK or B. Braun Melsungen AG, FR; pH 5.5) and the same dialysis fluids adjusted to pH 7.5 (with NaOH) were tested for the presence of a chemoattractant. Peritoneal effluents of CAPD patients from an overnight exchange of at least 8 h were centrifuged (300 g, 10 min). The supernatants, cell-free PE, were tested without concentration or dilution for chemotactic activity in the chemotaxis assay. For a doseresponse curve, the PE was first concentrated SO-fold with YM2 (rI kDa) Diaflo Ultrafilters (Amicon Corp., Scientific Systems Division, Denvers, MA, USA) before testing the different dilutions in RPMI 1640 for chemotactic activity. Before fractionation with high-pressure liquid chromatography (HPLC) the PE was first concentrated 50-fold with YM2 ( > 1000 Da) Diaflo Ultrafilters. In the fractions with a molecular weight less than 1 kDa no chemotactic activity could be determined. Some PE were digested with trypsine-conju-

gated polyacrylamide beads, before fractionation [lo]. The 50-fold-concentrated PE were fractionated by HPLC size exclusion in a TSK G 3000 SW column (LKB, Bromma, Sweden; 7.5 x 600 mm) with a separation range of 1-3000 kDa. Fractions were tested in the lower compartment of the chemotaxis chamber in the same dilution as the starting material. Agents used in rhe migration assuy

As commercial agents we used the synthetic chemoattractant N-formyl-methionyl-leucyl-phenyl-alaninemethyl-ester (FMLP; Sigma Chemical Company, St Louis, MO, USA), Casein nach Hammerstein (Merck. Darmstadt, FRG), recombinant interleukin-I a (rlLl a ; specific activity: 1 x lo7 U [lymphocyte activation factor; [LAF] ml-’; [ I I]) kindly donated by Roche. Herfordshire, UK, and lipopolysaccharide (LPS; lipopolysaccharide B; B 01 11: B4 from E. coli, Difco Laboratories, Michigan, ND, USA). All chemoattractants were diluted in RPMI 1640 medium before testing in the chemotaxis assay. The detection ofconiplt~tnetit~ucr~r 5u unil inttvdeukiri I in the peritonrul efluent

The detection of the presence of complement factor 5a (C5a) in the PE was tested with a radioimmunoassay kit (Uppjohn Diagnostic, Kalamazoo. MI, USA) and interleukin 1 (IL-I) was assayed using the thymocyte proliferation assay [ I I]. The ckemotuxis assay Chemotaxis was performed in a 48-well microchemotaxis chamber [12]. Twenty-five p1 of different solutions were placed in the wells of the bottom chamber. A filter sheet (Nucleopore Incl., Cabin John, MD, USA), polyvinylpyrrolidone-free (PVP-F), I0 pm thick, 8 pm pore size, was placed over the wells. A PVPF polycarbonate filter was used to prevent migrated PC from dropping off the lower filter surface [13]. Gasket and top plate were assembled and the apparatus warmed up to 37 C. PC adjusted to 20000 in 50 111 a-RPMI (unless indicated otherwise), were added to each top well and the chambers were incubated in moist air with 5% CO? at 37 C for 3 h. After incubation, the chamber was disassembled and the filter was removed. Non-migrated cells which remained on the top side were wiped off and the filter was air-dried, fixed in methanol and stained in Brilliant Blue R (Sigma). For each well, 20 areas were counted (a total filter area of 0.432 mm’). The average was used to determine the number of migrating cells in the total available area (8 mm’)). The values were expressed as the percentage of cells originally loaded in the top chamber that migrated through the filters. Each test was performed in triplicate and the percentage of migrated cells was expressed as the mean. To test the

CHEMOTAXIS IN CAPD PATIENTS presence of a chemoattractant in the PE, we performed a 'checkerboard assay' [14]. Results Churuc,trri=utioriof the prritontwl cells

The average cell recovery from the total PE of CAPD patients and healthy volunteers was 12.6 x lo6 respectively (viability >9O'X,). The PC from the CAPD patients consisted on average of 83% macrophages, 13% lymphocytes, 2%) granulocytes, 1 '%,eosinophils and less than 1 %, mast cells. PC of healthy volunteers consisted on average of 90% macrophages and 10% other cell types, which were mainly lymphocytes. Migrution of perironed ce1l.s to FMLP, casein. rIL-la and LPS

In Fig. I , for every examined agent, dose-response curves with PC derived from five CAPD patients are shown. Almost all of the migrated PC were macrophages, as determined after studying May-Griinwald-

Giemsa-stained polycarbonate filters. Also all migrated cells adhered to the polycarbonate filter, since in the bottom wells no cells could be detected. As shown in Fig. la an optimal migration of PC to FMLP mol I - ' . Casein occurred at a concentration of gave a response over a broad concentration range, but an optimum occurred at a concentration of 1 mg ml-' (Fig. Ib). PC responded to rIL-la and a maximum occurred at a concentration of 1000 U (LAF) ml-' (Fig. Ic). PC also migrated to endotoxin (LPS) with a maximum response at 1 pg ml-' (Fig. Id). As shown in Fig. 2 , PC from CAPD patients showed a varied specific response to the optimal concentration of FMLP mol I-'). while the response of PC from healthy volunteers to FM LP was significant lower (Wilcoxon test, P