Isolation of rat peritoneal mononuclear and polymorphonuclear leucocytes on discontinuous gradients of Nycodenz

Journal of Immunological Methods, 133 (1990) 31-38 Elsevier

31

JIM 05688

Isolation of rat peritoneal mononuclear and polymorphonuclear leucocytes on discontinuous gradients of Nycodenz Sanne Fisker, Kirsten Kudahl and Ole Sonne Institute of Physiology, University of Aarhus, Aarhus, Denmark

(Received 5 February 1990, revised received 11 April 1990, accepted 29 May 1990)

Centrifugation of rat leucocytes from thioglycollate-induced inflammatory peritoneal exudate on a discontinuous gradient of Nycodenz with a density of 1106 g/1 and an osmolarity of 400 mosmol/1 separated the polymorphonuclear from the mononuclear leucocytes. The cells on the interphase between the buffer and the gradient medium contained 96% mononuclear leucocytes with a recovery of > 60%, and the bottom fraction consisted of 98% polymorphonuclear leucocytes with a yield of 91%, when an exudate isolated 20 h after the injection of thioglycollate was fractionated. Leucocytes isolated from non-inflamed rat peritoneum could be enriched in the fraction of nonspecific esterase-positive cells from 86% to 96% with a recovery of 82% on a gradient with a density of 1091 g/1 and an osmolarity of 325 mosmol/1. The viability of the isolated cells was > 95% (trypan blue exclusion tes0, and there was no measurable reduction in the fraction of phagocytosing cells (latex and opsonized zymosan) after exposure to the hypertonic gradient material. Key words: Nycodenz; Gradient centrifugation; Cell separation; Macrophage, peritoneal; Mononuclear leukocyte; Polymorphonuclear leukocyte; Phagocytosis; Nonspecific esterase; (Rat)

Introduction

The mononuclear (MNL) and polymorphonuclear (PMN) leucocytes from various animal species possess characteristics differing from those of the corresponding human cell types. Therefore, the isolation of these cell types from animals cannot follow the procedures using commercially available ready-to-use centrifugation media which

Correspondence to: O. Sonne, Institute of Physiology, University of Aarhus, Ole Worm's All* 160, DK-8000 Aarhus C, Denmark. Abbreviations: Hepes, 4-(2-hydroxyethyl)-l-piperazineethenesulphonic acid; MNL, mononuclear leucocytes; PMN, polymorphonuclear leucocytes.

are based on the sedimentation characteristics of the human blood cells (Boyum, 1976). The initial response to the induction of a sterile inflammation in the pleura1 or peritoneal cavities is an exudate enriched in PMN (within 24 h), followed by an increasing concentration of MNL. 3-4 days after the induction of the inflammation, the cell population is mainly accounted for by MNL (Hurley et ai., 1966). It is common to harvest PMN 4-24 h after the injection of a stimulant (e.g., Bray et al., 1987; Hornebeck et al., 1987; Lichtenstein, 1987) and to harvest MNL 3-4 days after the injection (e.g., Kaplan et al., 1979). Usually no further purification is employed although PMN fractions may contain up to 15% MNL (Bray et al., 1987). In some studies, MNL have been purified by adherence during a subsequent tissue culture step (e.g., Kaplan, 1977).

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

32

We wished to study certain cellular parameters as a function of the time after induction of inflammation, and consequently required pure cell populations. For this purpose, we designed density gradients using the nonionic triiodinated derivative of benzoic acid, Nycodenz, as the separation medium. Furthermore, we took advantage of differences in the buoyant density of the two cell types in response to variations of the osmolarity of the extracellular medium (B~yum, 1968). By using a hyperosmotic gradient medium, it was possible to increase the difference in buoyant densities, so that a better separation was achieved. Here we report the composition of a gradient giving highly purified populations of both PMN and MNL from the same sample, and with minimal loss of cells.

Materials and methods

Animals Male Wistar rats fed ad libitum and weighing 190-230 g were used throughout this study.

Isolation of cells Rats were killed by cervical dislocation, and the leucocytes collected from the peritoneal cavity by lavage with 40 ml of icecold Krebs-Ringer salt solution buffered to pH 7.4 with 25 mmol/1 4-(2hydroxyethyl)-l-piperazineethanesulphonic acid (Hepes, Boehringer, Mannheim, F.R.G.) and supplemented with 10 g/1 of dialysed bovine serum albumin (Sigma, St. Louis, MO, U.S.A.) (Sonne, 1985). The cells were centrifuged at 20 °C for 10 min at 300 × gmax, and washed once in the same buffer. When indicated, the rats received an intraperitoneal injection of 3 ml of NIH thioglycollate broth (0.26 g/ml) (Difco, Detroit, MI, U.S.A.) 1-4 days prior to the intraperitoneal lavage.

Gradient centrifugation The density of the Nycodenz gradient material (Nycomed, Oslo, Norway) was 1150 g/l, and the osmolarity was 290 mosmol/1. This was made up to the specified densities and osmolarities by mixing with a hyperosmolar NaCl-based diluent (cf. legend to Table I) and water (Boyum et al., 1983).

TABLE I NYCODENZ GRADIENTS Gradient A was used for separation of peritoneal exudate cells from stimulated rats, and gradient B was used for enrichment of nonspecific esterase-positive peritoneal cells from unstimulated rats. Density Osmolarity Volume fractions of (g/l) A 1106 B 1091

(mosmol/l) Nycodenz a Hyperosmolar Water solution b 400 325

0.7006 0.5986

0.0788 0.0606

0.2206 0.3408

a Nycodenz, analytical grade, sterile solution from the manufacturer (1150 g/l, 290 mosmol/l). b 5.0 mmol/l Hepes, 3.0 retool/1 KC1, 0.3 mmol/l Na2EDTA , 0.3 mmol/1 CaCI2, and 1245 mmol/l NaC1 (1003 g/l, 2.5 osmol/1).

The compositions of the recommended gradients are shown in Table I. Unless otherwise stated, a maximum of 7 ml of leucocyte suspension with 6 × 10 7 cells were transferred to 10 ml tubes, and 3 ml of gradient medium (20 °C) were underlayered with a long needle attached to a syringe. After centrifugation at 700 × gmax for 20 min at 20 °C in a swing-out rotor, the cells at the interface (the top fraction) were aspirated with a pasteur pipette. The supernatant and the gradient medium were aspirated and discarded. The top and bottom fractions of cells were washed once as described above. Platelets found in the top fraction were removed by the washing procedure. There were only a few or no erythrocytes in the withdrawn peritoneal fluid, and they did not interfere with either the counting or the separation of the cells.

Identification of cells The cells were counted in a Btirker-Ti~rk haemacytometer. The viability of the recovered cells was judged by the Trypan blue exclusion test. The differentiation between PMN and MNL was based on microscopy of leucocytes stained with 9 vol. of 0.5 g/1 of methyl violet (Merck, Darmstadt, F.R.G., C.I. no. 42535) in 90 nunol/1 acetic acid. In contrast to human cells, polymorphonuclear and mononuclear peritoneal leucocytes from the rat could not be distinguished at a magnification of 400 times. Consequently this was

33 done using an oil immersion lens (1000 times final magnification). Nonspecific esterase assay. 2.5 mg et-naphthyl acetate (Sigma) were dissolved in 0.5 ml methanol and mixed with 12.5 ml of a sodium phosphate buffer (50 mmol/1, pH 7.8). 12.5 mg Fast Red RC diazotized salt (Sigma) were added under magnetic stirring and the solution filtered through Whatman 42 filter paper (Gomori, 1957). A sample of the cell suspension and the staining solution were mixed in the proportion 1 : 5 and incubated for 20 rain at 20 °C (Ennist and Jones, 1983). The macrophages and monocytes appeared red-brown. All solutions, except for the sodium phosphate buffer, were freshly made.

Assays for cellular integrity Phagocytosis of latex. This was performed by mixing 10 #1 of a latex bead suspension (Latex 0.81, Difco) with 5 × 105 cells in 1 ml, followed by centrifugation at 1000 x gmax for 10 min. After 2 h of incubation at 37 ° C, the cell suspension was washed three times in 2 ml buffer. A specimen of the cell suspension was stained with methyl violet, and the fraction of cells that had endocytosed at least two particles was determined by counting > 100 cells.

Phagocytosis of opsonized zymosan. Zymosan (Sigma) was dissolved in 15 mmol/1 Tris-HC1, 150 mmol/1 NaC1, pH 7.4, to a final concentration of 4 g/1 and was opsonized on the day of the analysis for 30 min at 37°C in 20% freshly drawn serum or heat-inactivated serum (Hallrn-Sandgren and Bj~Srk, 1988). The MNL were incubated with the zymosan particles for 30 min at 37 ° C, and phagocytosis quantified as described above.

Receptor binding of 1251-labelled a2-macroglobulin-trypsin complex. This was performed as previously described (Sonne, 1988). In brief, the cells (3 × 109/1) were incubated with 10 pmol/1 x25I-labelled a2-macroglobulin, trypsin complex for 20 h in a shaking water bath at 4 ° C. Human et2-macroglobulin (a gift from Dr. Lars Sottrup-Jensen, Institute of Molecular Biology, University of Aarhus) was prepared from outdated, pooled citrated plasma using Zn2+-chelate affinity chromatography as previously described in detail (Sottrup-Jensen et al.,

1978). It was iodinated with 1251 to a specific activity of about 6.3 x 1016 Bq/mol using chloramine-T as the oxidizing agent, complexed with trypsin as previously described (Gliemann and Davidsen, 1986), and purified on a column of Sephacryl S-300 (Pharmacia, Uppsala, Sweden). The trapped tracer was assessed by the addition of unlabelled ct2-macroglobulin, trypsin complex to give a final concentration of 110 nmol/1. This value averaged 3% of the total binding and was subtracted from the data presented. The incubations were stopped by centrifugation of the cells from a 200 /~l aliquot of the incubation mixture through a layer of silicone oil (a 1 : 2, v/v, mixture of AR20/AR200 from Wacker-Chemie, Munich, F.R.G.) in a Beckman Microfuge. The cell pellet was isolated by cutting the tube through the oil layer (Gliemann and Davidsen, 1986).

Results

Separation of peritoneal exudate MNL and PMN from thioglycollate-stimulated rats When the rats received an intraperitoneal injection of thioglycollate 20 h prior to the harvest of the cells, the peritoneal exudate cells consisted of about 60% PMN and 40% MNL (Table II). This mixed cell population was separated by discontinuous gradient centrifugation as described in the materials and methods section. All centrifugations resulted in transparent gradient media, indicating a complete separation of the top and bottom fractions. The top fraction consisted mainly of MNL, while the PMN were mainly found in the bottom fraction. The PMN change their volume and their density more than MNL in response to a hyperosmotic medium (Boyum, 1968). The effect of varying the osmolarity at different densities of the gradient medium was, therefore, investigated in an attempt to optimize the purity and the recovery of the t w o cell types. As the osmolarity was increased, the purity of the MNL in the top fraction increased, but there was a concomitant decrease in the recovery of MNL (Fig. 1A). Although this phenomenon persisted at higher densities, it could to some extent be counteracted by increasing the

34 T A B L E II PERITONEAL EXUDATE CELLS O B T A I N E D AT VARIOUS INTERVALS AFTER I N T R A P E R I T O N E A L I N J E C T I O N OF THIOGLYCOLLATE Methyl violet differential counting was undertaken before and after centrifugation on a gradient of Nycodenz with a density of 1106 g / l , and an osmolarity of 400 mosmol/1 as described in the materials and methods section. M e a n + S D of 2 - 4 independent experiments. Days after injection 1

2

3

4

N u m b e r of experiments (3)

(2)

(3)

(4)

MNL* PMN a

40+ 4 60+ 4

71+20 29+20

86+ 3 14+ 3

98+ 2 2+ 2

Purity

MNLtopb PMNbottom c

96+ 1 98+ 1

95+ 3 91+ 5

99+ 1 82+ 8

99+ 1 -

Recovery

M N L top d P M N bottom e

60+17 91 + 23

73+ 8 81 + 18

66+14 56 + 22

75+27 -

Before centrifugation

After centrifugation

a b c d e

The The The The The

fraction of M N L a n d PMN, respectively, in the starting material expressed as per cent of the total n u m b e r of cells. fraction of M N L in the top fraction after centrifugation expressed as per cent of the total n u m b e r of ceils in this fraction. fraction of P M N in the bottom fraction after centrifugation expressed as per cent of the total n u m b e r of cells in this fraction. recovery of the M N L in the top fraction expressed as per cent of the total n u m b e r of M N L in the starting material. recovery of the P M N in the bottom fraction expressed as per cent of the total n u m b e r of P M N in the starting material.

density from 1.086 g/1 to 1096 and 1106 g/1 (Figs. 1B and 1C versus 1A). With increasing osmolarity, the purity of the PMN in the bottom fraction decreased, while the yield increased (Fig. 2A). When both the osmolarity and the density were increased, a high purity as well as a high yield of PMN were obtained (Fig.

2c). The combination of the results in Figs. 1C and 2C reveals that an optimal separation was achieved at a density of 1106 g/1 and an osmolarity of 400 mosmol/1 (A in Table I). By increasing the osmolaxity to 420 mosmol/1, there was a slight increase in the purity of the MNL (Fig. 1C), while the purity of the PMN (Fig. 2C), and the recovery of cells in both fractions decreased (Figs. 1C arid 2C). The above results were obtained with cells isolated 20 h after intraperitoneal injection of thioglycollate. At this time, the PMN constitute approximately half of the total number of cells (Table II). In order to investigate the ability of the gradient to purify the two cell types, when their proportions differ in the starting material, peri-

toneal exudate cells were isolated at various intervals after the intraperitoneal injection of thioglycollate. On the first day after the injection separation was optimal because of the great excess of PMN in the exudate. However, when the fraction of MNL was dominant (days 2 and 3), this cell type increasingly contaminated the PMN in the bottom fraction. On the other hand, the purity of the MNL in the top fraction increased, as the fraction of PMN in the starting material decreased (Table (II). If too many cells are loaded onto the gradient, cells trapped at the interface between the buffer and the gradient medium will constitute a mechanical barrier hindering denser cells from penetrating the layer of buoyant cells. Furthermore, when the cells are exposed to the hyperosmotic gradient media, water is expelled from the cytoplasm. This will decrease both the osmolarity and the density of the gradient medium. There is therefore an upper loading limit on each gradient. We found that a maximum of 6 × 107 cells could be loaded on gradients in tubes with an inner diameter of 13

35

A

8% (n = 6) esterase-negative cdlls. To isolate the esterase-positive cells from this material, centrifugation on a gradient with a density of 1091 g/1 and an osmolarity of 325 mosmol/1 (B in Table I) gave a purity of 96 + 3% and a recovery of 82 + 9% (n = 6) of esterase-positive cells in the top fraction (data not shown).

-100

-80 100-

~

-60 -40

8060

I'-. ~

,

,

-50

,

,

,

~

0 "100

B

-80

~100. t~

~ "~-----" T -----.T _

s060

~5o ~

,

,

~

,

,

~

0 "100

C

T A 100-

......

T I . . . . ~. . . .

~

8060

,-"

~so ~

.80 ~s0 -40

-20 ~

500 3~o 350 3;0 3~o 3'-o ,~o 4~o . o

0

Osmolarity (naOsrnol/l) Fig. 1. The effect of the density and the osmolarity of the gradient medium on the purity and yield of M N L in the top fraction. A mixed peritoneal exudate cell population consisting of M N L and P M N was centrifuged on discontinuous gradients with differing densities and osmolarities. The top fraction was isolated and stained with methyl violet.The purity (e e) was calculated as the fraction of M N L in the total number of leucocytes present in the top fraction. The yield (zx zx) was defined as the number of recovered M N L in the top fraction divided by the number of M N L present in the starting material. The results presented here arc made for the purposes of presentation, since the development of the combinations of osmolarity and density of the gradients were established on the basis of several other trials,some of which also exceeded the presented limits. Bars represent S D of three independent experiments, when these exceed the size of the symbol. The densities of the gradient media were A: 1086 g/l; B: 1096 g/l; C: 1106 g/l.

Integrity of the isolated cells The viability of cells recovered from the fractions was always > 95% as judged by the trypan blue exclusion test (data not shown~r. In order to investigate the influence of the gradient medium on the functional integrity of the cells, leucocytes from unstimulated rat peritoneum exposed for 20 min at room temperature to the gradient medium with an osmolarity of 400 mosmol/1 and a density of 1106 g/1 were compared with leucocytes exposed to isotonic buffer under otherwise identical conditions. There was no difference in the phagocytosis of latex or of opsonized zymosan, or in the

2~

-80

100-

- 40 00-

-20

,

00

B

Isolation of esterase-positive MNL from the unstimulated rat peritoneal cavity The MNL isolated 1-4 days after injection were mainly of the monocyte/macrophage lineage as indicated by the presence of 92-98% esterasepositive cells. By contrast, leucocytes collected from unstimulated rat peritoneum contained 14 +

0 -lOO -80 -8o

lO0-

- 40

,~

SO-20

60

mm (i.e., 4.5 × 107 cells/cm 2) without interfering with the recovery of both cell types or with the purity of the two fractions (Table III). At higher loadings the gradient media became cloudy and there was a decreased recovery (data not shown).

-100

-80

i

i

,

i

/I .... i .....i c

0 -100

-80 -80

100-

-40 80-

60580

-20 300

3~o

3;0 3~o ,;o ,~o . o

0

Osmolarity (mOsmol/1) Fig. 2. The effect of density and osmolarity of the gradient media on the purity and yield of PMN in the bottom fraction. The purity (It It) was calculated as the fraction of PMN in the total number of leucocytes present in the bottom fraction. The yield (A 4) was defined as the number of recovered PMN in the bottom fraction divided by the number of PMN present in the starting material. Other conditions are as described in the legend to Fig. 1. The results are obtained from the same gradients as shown in Fig. 1.

36 TABLE III RECOVERY A N D PURITY O F THE CELL FRACTIONS AS A F U N C T I O N O F THE LOAD ON THE G R A D I E N T Cells in 7 ml obtained by peritoneal lavage consisting of 42% M N L and 58% PMN were subjected to centrifugation on 3 ml of a Nycodenz gradient with a density of 1106 g / l and an osmolarity of 400 mosmol,/l in a tube with an internal diameter of 13 mm as described in the materials and methods section except that the load was varied as indicated in the table. Mean 5: SD of three independent experiments. Load( × 107 cells/tube) 6

9

12

98+2 98+2

99+ 0 99+ 1

98+1 98+0

69+1 104+8

40+11 98+ 7

51+2 91+4

Purity MNL in top a PMN in bottom b

Recovery MNL in top e PMN in bottom d

a The fraction of MNL in the top fraction after centrifugation expressed as per cent of the total number of cells in this fraction. b The fraction of PMN in the bottom fraction after centrifugation expressed as per cent of the total number of cells in this fraction. c The recovery of the M N L in the top fraction expressed as per cent of the total number of MNL in the starting material. d The recovery of the PMN in the bottom fraction expressed as per cent of the total number of PMN in the starting material.

require other specifications of the gradients in order to be optimal. Hopper et al. (1985) isolated murine PMN from 20 h post-injection exudate to 98% purity on a discontinuous Metrizamide (Nycomed) gradient, and MNL to 92% purity on discontinuous Percoll gradients with the majority of the lymphocytes separated into other fractions than the macrophages. In this study no information was provided on the recoveries obtained. In another study using self-generating Percoll gradients, murine PMN were purified to 78% with a recovery of 35-50% from peritoneal exudate isolated 24 h after an injection of Corynebacterium parvum (Lichtenstein et al., 1985). In the same study MNL were purified to 97% with a recovery of 45-60%. This is one of the few reports in which both PMN and MNL

TABLE IV T H E F U N C T I O N A L I N T E G R I T Y OF M O N O N U C L E A R LEUCOCYTES EXPOSED TO HYPEROSMOLAR G R A D I ENT MATERIAL Resident peritoneal M N L were isolated by peritoneal lavage and exposed to a Nycodenz solution with a density of 1106 g / l and an osmolarity of 400 mosmol/1 (A in Table I) or to isotonic buffer for 20 min at room temperature. After a wash, the cells were analysed for their ability to phagoeytose latex and opsonized zymosan, and to bind a2-macroglobulin-trypsin complex. Mean + SD of independent experiments. n Cells exposed to

receptor binding of a2-macroglobulin-trypsin complexes between gradient-exposed and control cells (Table IV).

Buffer

Gradient medium

Phagocytosis Discussion

Although the procedure described by Boyum (1976) using sodium metrizoate-Fico11 for the isolation of MNL and PMN to homogeneity has set the standard for human blood, other gradient media such as Nycodenz (Nycomed) and Percoll (Pharmacia) have also been used (Ford et al., 1987; Davies and Lloyd, 1989). However, since the commercially available mixtures (e.g., Lymphoprep from Nycomed) are optimized for human blood cells, the different sedimentation characteristics of these cell types from other species will

Latex a 3 84 + 1 Zymosan opsonized with serum b 4 77+5 heat-inaetivatedserum b 4 1 6 + 8

80 + 5 77+3 19+3

Receptor binding a 2-macroglobulin-trypsin c 3

0.023+0.003

0.024-t-0.008

a Cells having endocytosed at least 2 latex particles during the phagocytosis assay described in the materials and methods section expressed as per cent of the total number of cells. b Cells having endocytosed at least two zymosan particles during the phagocytosis assay described in the materials and methods section expressed as per cent of the total number of cells. ¢ The binding of 125I-labeiled a2-macroglobulin-trypsin complex was assayed as described in the materials and methods section and expressed as the B/F ratio per 109 cells//1.

37 were obtained from the same gradient. However, since the polyvinyl-pyrrolidone-coated colloidal silica-gel particles constituting the centrifugation medium Percoll has been reported to interfere with the function of phagocytosing cells (Wakefield et al., 1982), we have preferred to use the iodinated density gradient medium, Nycodenz. Centrifugation on a gradient medium with a density of 1106 g/1 and an osmolarity of 400 mosmol/1 of peritoneal lavage cells from rats having received an injection of thioglycollate gave both M N L and P M N at a high purity (96 and 98%, respectively) as well as a high yield (60 and 91%, respectively). The advantage of further increasing the osmolarity to 420 mosmol/1 was negligible. The highest possible purity is always desirable, and a high yield will minimize the number of animals needed for an experiment. However, a high yield is also of importance in order to ensure that the cell fraction under investigation is representative of the whole population. 2 and 3 days after the thioglycollate injection, the M N L outnumbered the PMN, and it was no longer possible to obtain the latter fraction with high purity (91% at day 2, and 82% at day 3; Table II). Despite this, the gradient described in this paper is among the most versatile yet described in achieving highly purified cell fractions with a relatively high recovery (Table I). H u m a n P M N reportedly maintain unaltered phagocytic activity, bactericidal capacity and chemotaxis after exposure to media of osmolarities varying from 200 to 600 mosmol/1 (Dooley and Takahashi, 1981). In this study the M N L exposed to gradient material with an osmolarity of 400 mosmol/1 showed normal phagocytosis of both latex and opsonized zymosan (Table IV). Furthermore, the specific receptor binding of 1251labelled a2-macroglobulin, trypsin complex was also normal (Table IV). Dead human monocytederived macrophages are unable to bind a2-macroglobulin-trypsin (Petersen et al., 1987), and in rat hepatocytes the cellular concentration of this receptor has been shown to decrease dramatically even after short incubation periods at 3 7 ° C (Sonne, 1988). Thus, the a2-macroglobulin.protease complex receptor appears to be fragile in in

vitro systems and is, therefore, a sensitive index of the integrity of cells expressing this receptor.

Acknowledgements S.F. and K.K. were the recipients of scholarships from K o b m a n d i Odense Johann og H a n n e Weimann, F. Seedorffs Legat, the D a n i s h Ministry of Education (Biotechnology scholarships), and Fonden til Laegevidenskabens Fremme. The first two authors contributed equally to this study. This study was in part supported by grants from the N o v o Foundation, the Danish Biomembrane Research Centre, and Konsul Johannes Fogh-Nielsens og fru Ella Fogh-Nielsens Legat.

References Boyum, A. (1968) A one-stage procedure for isolation of granulocytes and lymphocytes from human blood. General sedimentation properties of white blood cells in a 1 g gravity field. Scand. J. Clin. Lab. Invest. 21 (suppl. 97), 51. Boyum, A. (1976) Isolation of lymphocytes, granulocytes and macrophages. Scand. J. Immunol. 5 (suppl. 5), 9. Boyum, A., Berg, T. and Blomhoff, R. (1983) Fractionation of mammalian cells. In: D. Rickwood (Ed.), Iodinated Density Gradient Media - A Practical Approach. IRL Press, Oxford, p. 147. Bray, M.A., McKearn-Smith, C., Metz-Virca, G., Bodmer, J.L. and Virca, G.D. (1987) Stimulated release of neutral proteinases elastase and cathepsin G from inflammatory rat polymorphonuclear leukocytes. Inflammation 11, 23. Davies, D.E. and Lloyd, J.B. (1989) Monocyte-to-macrophage transition in vitro. A systematic study using human cells isolated by fractionation on Percoll. J. Immunol. Methods 118, 9. Dooley, D.C. and Takahashi, T. (1981) The effect of osmotic stress on the function of the human granulocyte. Exp. Hematol. 9, 731. Ennist, D.L. and Jones, K.H. (1983) Rapid method for identification of macrophages in suspension by acid alpha-naphthyl acetate esterase activity. J. Histochem. Cytochem. 31, 960. Ford, T.C., Needle, R. and Rickwood, D. (1987) A rapid method for the preparation of the neutrophil fraction of granulocytes from human blood by centrifugation on isotonic Nycodenz gradients. Blut 54, 337. Gliemann, J. and Davidsen, O. (1986) Characterization of receptors for a2-macroglobulin-trypsin complex in rat hepatocytes. Biochim. Biophys. Acta 885, 49. Gomori, G. (1957) Histochemical methods for enzymes. Methods Enzymol. 4, 381.

38 Halldn-Sandgren, C. and Bjtrk, 1. (1988) A rapid technique for the isolation of highly purified, functionally intact bovine neutrophilic granulocytes. Vet. Immunol. Inununopathol. 18, 81. Hopper, K., Hollister, W., Davey, R. and Scmler, A. (1985) Release of galactosyltransferase from peritoneal macrophages during acute inflammation. J. Cell. Physiol. 124, 137. Homebeck, W., Soleillmc, J.M., Tixier, J.M., Moczar, E. and Robert, L. (1987) Inhibition by elastase inhibitors of the formyl Met Leu Phe-induced chemotaxis of rat polymorphonuclear leukocytes. Cell Biochem. Funct. 5, 113. Hurley, J.V., Ryan, G.B. and Friedman, A. (1966) The mononuclear response to intrapleural injection in the rat. J. Pathol. Bacteriol. 91, 575. Kaplan, G. (1977) Differences in the mode of phagocytosis with Fc and C3 receptors in macrophages. Stand. J. Immunol. 6, 797. Kaplan, G., Unkeless, J.C. and Cohn, Z.A. (1979) Insertion and turnover of macrophage plasma membrane proteins. Proc. Natl. Acad. Sci. U.S.A. 76, 3824. Lichtenstein, A. (1987) Stimulation of the respiratory burst of murine peritoneal inflammatory nentrophils by conjugation with tumor cells. Cancer Res. 47, 2211.

Lichtenstein, A., Kalde, J. and Bonavida, B. (1985) Use of a self-generating Percoll gradient and single cell cytotoxicity assay to identify tumor-lyric properties of inflammatory neutrophils. J. Immunol. Methods 81, 95. Petersen, C.M., Ejlersen, E., Hansen, P.W. and Gfiemann, J. (1987) Binding of alpha-2-macroglobulin trypsin complex to human monocytes in culture. Scand. J. Clin. Lab. Invest. 47, 55. Sonne, O. (1985) In vitro biological characterization of iodinated insulin preparations. In: J. Lamer and S.L. Pohl (Eds.), Methods in Diabetes Research, Vol. I, part C: Laboratory Methods. John Wiley, New York, p. 433. Sonne, O. (1988) The cellular dynamics of hepatic receptors for a2-macroglobulin-protease complex and for insulin are different. J. Biochem. (Tokyo) 103, 348. Sottrup-Jensen, L., Stepanik, T.M., Wierzbicki, D.M., Jones, C.M., I_~nblad, P.B., Kristensen, T., Mortensen, S.B., Petersen, T.E. and Magnusson, S. (1983) The primary structure of a2-macroglobulin and localization of a Factor XIII a cross-linking site. Ann. N.Y. Acad. Sci. 421, 41. Wakefield, J.S.J., Gale, J.S., Berridge, M.V., Jordan, T.W. and Ford, H.C. (1982) Is Percoll innocuous to cells? Biochem. J. 202, 795.