Roles of antibody and complement in the bactericidal ... - Europe PMC

0 downloads 0 Views 1MB Size Report
and normal mouse (NMS) sera. Opsonization by IMS .... under anaesthesia, and the sera collected, filtered and stored as for ... The perito- neal cavities were ...
Br.

J. exp. Path. (I987) 68, 89-Ioo

Roles of antibody and complement in the bactericidal activity of mouse peritoneal exudate neutrophils Prue H. Hart, Lyn K. Spencer, Nanette L. Hill, Peter J. McDonald and John J. Finlay-Jones Clinical Microbiology Unit, School of Medicine, Flinders University of South Australia, Bedford Park, S.A. 5042, Australia

Received for publication I 5 May I986 Accepted for publication 20 September I986

Summary. The contributions of complement and antibody to phagocytosis and, as a separate process, intracellular killing of Proteus mirabilis, were investigated using mouse peritoneal exudate neutrophils. Phagocytosis of P. mirabilis was promoted by both immune mouse (IMS) and normal mouse (NMS) sera. Opsonization by IMS promoted significantly greater phagocytosis than did NMS, as did NMS compared with heated IMS (HIMS). The ability of NMS to opsonize P. mirabilis for both phagocytosis and phagocytic killing was diminished by chelation with EGTA and abolished by chelation with EDTA. This suggested that fixation of complement by both alternative and classical pathways provided optimal opsonization of this organism in NMS. In order to study intracellular killing as a process separate from phagocytosis, peritoneal exudate cell suspensions were exposed to P. mirabilis, previously incubated with I% NMS, I% IMS, io% HNMS (heated normal mouse serum) or io% HIMS, followed by centrifugation of the phagocyte-bacteria mixtures on Percoll density gradients. Populations of neutrophils containing viable intracellular bacteria, and relatively free of extracellular bacteria (< 7% of total) were recovered in washed suspensions of cells fractionated at densities greater than I.o69 g/ml. For P. mirabilis that had been opsonized with i% NMS before phagocytosis, the continued presence of extracellular serum was necessary for intracellular killing. NMS stimulated significantly greater intracellular killing than did HNMS, which stimulated some intracellular killing compared with the absence of serum, in which no killing occurred. IMS was similar to NMS in its ability to stimulate intracellular killing. EGTA partially blocked the stimulation of intracellular killing by NMS, and EDTA abolished it. These findings suggested that (as for optimal opsonization) complement activated via both alternative and classical pathways was responsible for optimal stimulation of intracellular killing.

Keywords: neutrophil, phagocytosis, intracellular killing, serum factors, mice

Mouse neutrophils have receptors for both immunoglobulin and complement (Lopez et al. 198I). The roles that these receptors play

in the ingestion and killing of bacteria by murine neutrophils is not known. With respect to ingestion, it has been

89

P.H. Hart et al. found that the complement receptor is pri- essential for engulfment (Scribner & Fahrney marily involved in the attachment phase of 1976; Ehlenberger & Nussenzweig I977; opsonized red blood cells to mouse neutro- Hed & Stendahl I982), nor did we find the phils, and participation of the IgG receptor is extracellular presence of immunoglobulin to necessary for ingestion (Mantovani 1975). be as important for stimulation of intracelluSimilar results using opsonized red blood lar killing by murine neutrophils as others cells have been reported for mouse macro- found for human neutrophils (Leijh et al. phages (Ehlenberger & Nussenzweig I9 77; I98I) and mouse macrophages (Leijh et al. Griffin & Griffin 1979), and for human I984). neutrophils using opsonized red blood cells and opsonized Staphylococcus aureus Materials and methods (Scribner & Fahrney I976). With yeast, it was found that ingestion of IgG-coated orga- Mice. Male BALB/c mice, 8 to i 6 weeks old, nisms was significantly greater than inges- were used according to the ethical guidelines tion of C3b-coated organisms (Hed & Sten- of the National Health and Medical Research dahl I982). In contrast, we have found that Council and the Commonwealth Scientific phagocytosis and killing of the bacterium and Industrial Research Organisation of Proteus mirabilis by murine neutrophils Australia. requires an intact complement system in mouse serum, but will proceed in the absence Bacteria. The test organism, P. mirabilis, has of detectable specific antibody (Finlay-Jones been isolated from intra-abdominal abset al. I984), a finding of others with human cesses induced in mice by a complex inocuneutrophils and a number of strains of Gram- lum of mouse colonic and caecal contents negative bacteria (Leist-Welsh & Bjornson (Nulsen et al. I983). The organism was identified by the API 20E system (MontalieuI979). With respect to intracellular killing, as a Vercieu, France). Suspensions of log-phase process separate from ingestion, the conti- bacteria were prepared as previously denued presence of extracellular serum is scribed (Finlay-Jones et al. I984). Briefly, necessary for optimal intracellular killing of, small aliquots of frozen stocks of P. mirabilis in particular, catalase-positive organisms were grown aerobically overnight at 3 70C in such as staphylococci, by human neutro- trypticase soy broth (Difco, Detroit, MI, USA). phils (Leijh et al. 198I). Both complement Small samples were inoculated into fresh and immunoglobulins contributed to this broth for a further 3-4 h incubation. The stimulation, which was mediated through bacteria were harvested (I500 g, I5 min), cell-surface receptors. A similar requirement washed with o.9% saline and resuspended at exists for mouse peritoneal macrophages I x Io9/ml. The viable count was assessed (Leijh et al. I984). We have found that from pre-determined standard curves by extracellular serum is needed for optimal measuring absorbance at 420 nm. Viable intracellular killing of P. mirabilis by murine bacteria were enumerated according to their neutrophils (Hart et al. I985). ability to form colonies when appropriatelyWe have examined further the roles of diluted samples were spread onto CLED agar antibody and complement in the separate (Oxoid, Basingstoke, UK), and incubated processes of phagocytosis and intracellular aerobically overnight at 3 70C. killing of P. mirabilis by murine neutrophils. We found complement principally important Normal mouse serum (NMS). Blood was within both the opsonization of P. mirabilis for drawn from the retro-orbital venous plexus engulfment, and in the stimulation of intra- of anaesthetized mice. After clotting, the sera cellular killing. In contrast to the results of were pooled, filtered, and stored at - 70°C. others, we did not find immunoglobulin Heated normal mouse serum (HNMS) was

go

Stimulation of mouse neutrophil activity 9I prepared by incubation of NMS at 56°C for Peritoneal exudate cell collection. Mice were 30 min.

Immune sera. Mice were injected intraperitoneally (i.p.) with I X IO7 P. mirabilis or with I X IO7 E. coli in a volume of o.i ml on four occasions over a period of 3 months. Ten days following the last injection, the mice were bled via the retro-orbital venous plexus under anaesthesia, and the sera collected, filtered and stored as for NMS. Antibody titres were measured in an indirect immunofluorescence assay (Finlay-Jones et al. I984), using a fluorescein-conjugated goatanti-mouse IgG reagent (Nordic, Tilberg, The Netherlands). The P. mirabilis immune serum (termed IMS) had a titre of I/5 I 2, and the E. coli antiserum had a titre of I/1024. Antibody titres were unchanged over a storage period of I8 months. Heated immune serum to P. mirabilis was termed HIMS. Seven batches of NMS were tested by indirect immunofluorescence and had a mean titre of 1/1.3 ( range: not detected in undiluted serum-I/4). Chelation of serum. EDTA (BDH Chemicals, Australia) was prepared as a o. I M solution, with pH 7.2 obtained by addition of I M NaOH solution. MgCl2 (0.2 M) was combined with O.I M EGTA (Sigma Chemical Co., St Louis, MO, USA) to give a final solution of pH 7.2 achieved by addition of NaOH. Serum (o.9 ml aliquots) was incubated with EDTA or Mg2 +-EGTA solutions (o. i ml aliquots) for IO min at room temperature with occasional mixing before use in opsonization or intracellular killing experiments.

Opsonization of P. mirabilis. i x IO9 P. mirabilis, suspended in i ml RPMI-i64o medium supplemented with 20 mM HEPES, pH 7.2 (Flow Laboratories, McLean, VA, USA), were incubated with i ml serum (NMS, HNMS, IMS, HIMS, or chelated NMS) or i ml RPMII640 medium for 30 min in a 3 7°C reciprocating water bath. Bacteria were then washed twice (1500 g, 4°C, I5 min) with o.9% saline.

killed by cervical dislocation 3.5 h after i.p. injection of i ml Brain-Heart Infusion Broth (Oxoid) prepared with tap water. The peritoneal cavities were washed out initially with 5 ml, then with 3 ml, ice-cold mouse-osmolality phosphate-buffered saline-MPBS (Sheridan & Finlay-Jones 1977). The cells harvested were washed twice with MPBS (I 75 g, 7 min, 40C) before resuspension at 5 x 107 cells/ml in Hanks Balanced Salt Solution, Ca2 +- and Mg2 +-free (HBSS-Commonwealth Serum Laboratories, Melbourne, Australia), supplemented with I 0 mM HEPES, pH 7.2. Preparation of peritoneal exudate cells with intracellular bacteria. Peritoneal exudate cells (5 x IO7 cells/ml) were incubated with P. mirabilis (i x io8/ml) for 30 min in the presence of i% NMS or i% IMS, or for io min in the presence of IO% HNMS or io% HIMS in go x 13 mm polycarbonate tubes (Disposable Products, South Australia). The incubation times were selected in order to minimise intracellular killing of the organism during the period of phagocytosis. Incubations were carried out at 3 70C on an angled platform (Nutator: Clay-Adams, B-D and Co., Parsippany, NJ, USA) at 20 rotations/min. Phagocytosis was terminated by placing the tubes on ice and diluting the suspensions with icecold HBSS to the volume required for preparation of a 45% HBSS-diluted Percoll solution of density I.0575 g/ml. Samples (0.05 ml) were withdrawn at the beginning and end of each incubation for the determination of changes in the numbers of viable bacteria, and for the preparation of cytocentrifuge smears. For the former, samples were added to 5 ml O.I% Triton X-ioo (Ajax Chemicals, Sydney, Australia) in o.9% saline at room temperature for leucocyte lysis. After appropriate dilution in saline, o. i ml volumes were spread onto CLED agar and colonies counted after aerobic incubation overnight at 3 70C. Changes to the concentration of viable P. mirabilis were calculated according to the following formula:

P.H. Hart et al. 92 Assay of phagocytosis and determination of A logI 0(cfu/ml) = [logI 0(cfu/ml) at time t] [log1o (cfu/ml) at time o] leucocyte-associated and extracellular bacteria. 5 x 106 cells harvested from neutrophilThe cytocentrifuge smears were fixed with enriched bands following centrifugation on methanol before incubation with Jenner- Percoll density gradients were incubated Giemsa stain and subsequent determination with I X I07 P. mirabilis. In experiments of cell morphologies as well as the numbers comparing NMS with immune sera, the of internalized bacteria and the percentage of bacteria were added to suspensions of phagoneutrophils with them. cytic cells together with I0% serum. In experiments investigating the effects of Mg2+-EGTA or -EDTA chelation of NMS, P. Percoll solutions: Percoll (Pharmacia, Upp- mirabilis were first incubated with the approsala, Sweden) was diluted according to the priate serum for 30 min, washed twice with methods of Dooley et al. (I982). A stock saline and then incubated with phagocytic solution, labelled I00%, was prepared by cells in the absence of serum. Cell suspenmixing 9 volumes of Percoll with I volume of sions were prewarmed to 370C before the Ca2+- and Mg2+-free HBSS. Further addition of bacteria. Cell-bacteria suspenmixing with HEPES-buffered HBSS was persions were mixed on an angled platform formed to obtain Percoll solutions of 8I%, (Nutator: Clay-Adams) as 3 70C as described 70%, 55%, 5o%, and 45% concentration. above. Phagocytosis was terminated by placUsing the equation of Ulmer and Flad (I 9 7 9) ing the assay tubes on ice, and samples were these solutions were of density I.I002, taken for the determination of intracellular I.087I, I.0693, I.0634 and 1.0575 g/ml and extracellular bacteria. respectively. Distribution profiles of density For determination of leucocyte-associated marker beads (Pharmacia), after centrifugabacteria, samples of 0.05 ml were taken for tion on gradients of the Percoll solutions, preparation of cytocentrifuge smears, stainwere consistent with the calculated densities. ing, and examination by light microscopy. The term 'leucocyte-associated bacteria' has been used instead of 'phagocytosed bacteria' Density gradient centrifugation. Three ml of to denote that some of the organisms seen the most dense Percoll solution were disassociated with neutrophils may have been pensed into I00 I6 mm polycarbonate attached to the neutrophils, and not internatubes (Disposable Products, South Australized by them. For determination of extracellia). Without disturbance of the interface, 2 lular bacteria, we used the technique of ml of the 70% Percoll solution were added. Braconier and Odeberg (I979). NeutrophilSimilarly, 2 ml of each of the 55% and 50% bacterium mixtures were pulse-labelled with solutions were added. Finally, 3 ml of the 3H-thymidine, which does not penetrate 45% Percoll solution, in which were susphagocytes and is therefore unavailable to pended the cells for fractionation, were phagocytosed bacteria (Braconier & Odeberg layered on top of the gradient. The load was I979). Triplicate samples of 0.2 ml were approximately 5 x IO7 cells per gradient. withdrawn from assay mixtures, placed into The tubes were centrifuged at I 6oo g for 30 microtitre wells (Linbro, VA, USA) and 0.2 min at io'C. Cell bands were collected, in pCi 3H-thymidine (Amersham International, order, from the top of the gradients, using UK) in a volume of 0.002 ml added. After 30 pasteur pipettes. Gradient material between min at 3 70C, the well contents were harcell bands was discarded if not turbid. Cell vested with 5% trichloroacetic acid onto yields and per cent recovery were calculated glass fibre paper using a cell harvester (Titerafter two washes with MPBS (I 75 g, 7 min, tek, Skatron, Norway). After washing with 40C). water and methanol, the filters were dried at -

io x

x

Stimulation of mouse neutrophil activity 6o0C and the radioactivity determined using ACS II scintillation fluid (Amersham) and a Mark II beta scintillation counter (Searle Analytic Inc., IL, USA). The counts obtained were compared with similarly-processed control suspensions of bacteria without phagocytes. Bactericidal assay with ongoing phagocytosis. As previously described, 5 x IO5 P. mirabilis that had been opsonized in 50% NMS, HNMS, or chelated sera, were mixed with 5 x 106 peritoneal exudate cells in I ml of serum-free medium, and incubated with shaking at 3 7"C for 2 h (Finlay-Jones et al. I984). Samples were taken at o, i and 2 h for the enumeration of viable bacteria (intracellular plus extracellular). The leucocytes were lysed with o. i% Triton X-ioo, samples diluted with o.9% saline, and o. I ml volumes plated on CLED agar. Colonies were counted after overnight aerobic incubation at 3 70C. Assay of intracellular killing in the absence of ongoing phagocytosis (Hart et al. 1985). 5 x Io6 Percoll-fractionated, twice MPBSwashed cells with intracellular bacteria were combined with Io% (v/v) of the appropriate serum in a final volume of i ml RPMI. The 90 X 1 3 mm polycarbonate assay tubes were incubated at 3 70C on a rotating platform (Nutator: Clay-Adams) at 20 rotations/min. Samples (0.05 ml) were removed at zero time (i.e. when the tubes were removed from ice) and after a further 45 and go min for the enumeration of viable bacteria as above.

Expression of results. Mean values are given ± I s.d. The significance of differences between means was determined using Student's t-test. Results Fractionation of peritoneal exudate cells on discontinuous percoll gradients Centrifugation of peritoneal exudate cells on preformed discontinuous gradients of Percoll resulted in the separation of five well-differentiated bands of cells. The majority of

93

neutrophils were harvested in two bands: those above the Percoll solutions of densities I.O871 g/ml and I.1002 g/ml. Approximately equal numbers of neutrophils were isolated in these two fractions. Neutrophil purity in each band was usually in excess of 90%. Recovery of neutrophils in these two bands was approximately 70% of the neutrophils loaded onto the gradient (Hart et al. I985). For studies of phagocytosis, the cells banding between densities I.O87I g/ml and I.1002 g/ml were usually used. When peritoneal exudate cells were exposed to P. mirabilis for IO or 30 min before Percoll density gradient centrifugation, the density of the phagocytosing cells decreased and a greater proportion of the neutrophils were isolated at the interface between Percoll solutions of densities I.o8 71 g/ml and I *o693 g/ml. For studies of intracellular killing of P. mirabilis, cells harvested at densities greater than I.0693 g/ml were combined to yield viable, approximately 90% pure, neutrophil suspensions. Extracellular bacteria in these suspensions, as determined by their ability to incorporate 3H-thymidine, represented < 7% of the total bacteria present (Hart et al. I985).

Phagocytosis of P. mirabilis in the presence of NMS or IMS

Percoll-gradient-derivedneutrophils(5 x IO)

were incubated for IO min with I X 1O7 P. mirabilis in the presence of differing serum supplements at IO% final concentration. Phagocytosis, as determined by total leucocyte-associated bacteria, was significantly greater in IMS compared with NMS (Table i). Phagocytosis in HIMS was similar to that in NMS and significantly greater than that in HNMS, although the number of phagocytosing neutrophils in HIMS was less than in NMS. Heated antiserum to E. coli was similar to HNMS in its ability to support phagocytosis. The amount of radioactive precursor incorporated by the extracellular, non-phagocytosed bacteria was significantly less in IMS compared with NMS, and in NMS

P.H. Hart et al.

94

Table i. Phagocytosed and extracellular P. mirabilis in phagocytic assays incorporating different serum supplements Extracellular bacteriat

Phagocytosed bacteria*

Serum

Mean LAB/PMN Per cent PMN with LAB with LAB Total LAB P vs NMSt 3H-thy incorp Total ECB P vs NMSt I.2±0.2

io% NMS

44.3±6.4

1.7±0.I

32.0±0.7

IO% IMS

57.7±3.3

2-4±0.2

6i.8 i.I

io% HNMS

13-6±2.8 25.6±1.2 I5.5±o.6

3.0±o.8 2.3 ±0.3 2.4±0.2