Immunology 1976 31 263
The mediation of tissue eosinophilia in hypersensitivity reactions II. SEPARATION OF A DELAYED EOSINOPHIL CHEMOTACTIC FACTOR FROM MACROPHAGE CHEMOTACTIC FACTORS
M. HIRASHIMA, M. HONDA & H. HAYASHI Department of Pathology, Kumamoto University Medical School, Kumamoto, Japan
Received 22 December 1975; accepted for publication 2 February 1976
Summary. In anaphylactic cutaneous lesions induced by DNP-ascaris extract in the guinea-pig, the timecourse of delayed tissue eosinophilia was found to parallel that of the macrophage reaction, reaching its peak in 24 h. Macrophages could be differentiated I from lymphocytes by the numerous lysosomal I granules which stained for acid phosphatase. Extracts from such skin lesions contained a delayedI eosinophil chemotactic factor and two different macrophage chemotactic factors. Most of the delayed eosinophil chemotactic factor was separated from the two macrophage chemotactic factors by gel filtration on Sephadex G-100 and Sephadex G-200 in that order. The eosinophil chemotactic factor after re-chromatography on Sephadex G-200 showed no or little chemotactic activity for macrophages. INTRODUCTION The isolation and characterization of chemotactic factors present in inflammatory tissues is essential in order to clarify the mediators of inflammatory leucotaxis, though there remains the difficult Correspondence: Professor H. Hayashi, Department of Pathology, Kumamoto University Medical School, Kumamoto 860, Japan.
263
problem of obtaining sufficient amounts of chemotactic factors from tissues. As previously described (Hirashima and Hayashi, 1976), in active cutaneous anaphylaxis induced by DNP-ascaris extract in guinea-pigs, tissue eosinophilia manifested two phases; the early phase seemed to be associated with a thermostable factor with a molecular weight of less than 1400, probably dependent of mast cell degranulation and resembling chemotactic factors described by Parish and Coombs (1968) and Kay, Stechschulte and Austen (1971), while the delayed phase with a thermolabile factor with a molecular weight of about 70,000, probably independent of mast cell degranulation and difficult to compare with previously proposed chemotactic factors (see review by Parish, 1974). The delayed factor was considered to be more significant than the early factor, because the delayed tissue eosinophilia was more intense and prolonged than the early tissue eosinophilia. Since the chemotactic assay in the previous experiment was performed with eosinophils only, it remained to be resolved whether the delayed factor was selective for eosinophils. The purpose of the present communication is to describe a mainly parallel rise in the numbers of eosinophils and macrophages in the delayed phase, and a possible separation of the delayed factor from the macrophage chemotactic factors in the same skin extract.
264
M. Hirashima, M. Honda & H. Hayashi
MATERIALS AND METHODS Animals Male albino guinea-pigs of Hartley strain weighing 400 ± 50 g were used for induction of active cutaneous anaphylaxis by DNP-ascaris extract. The animals weighing 600-1000 g were used for separation of eosinophils from the peritoneal exudate and those weighing 300 ± 20 g for separation of peritoneal macrophages. Induction of active cutaneous anaphylaxis Immunization of animals was done by a modification of the method of Margni and Hajos (1973). The animals were injected intramuscularly in both hind legs with 0-5 ml of DNP-ascaris extract (DNP-As) (5 mg/ml in physiological saline) prepared by the method of Eisen, Belman and Carsten (1953) and Tada and Okumura (1971), with an equal volume of Freund's complete adjuvant (Difco, Detroit, Michigan) and accompanied by simultaneous intraperitoneal injection of 1-0 ml of killed B. pertussis (1 x 10'°/ml in physiological saline). Seven days later, the same injections were repeated. Induction of active cutaneous anaphylaxis was performed following the previous method (Hirashima and Hayashi, 1976). Twelve days after the first immunizing injection, the twenty immunized animals from the experimental group were intracutaneously injected with 0-1 ml of DNP-As (5 mg/ml in physiological saline) at eight sites over the body, and the treated sites were observed for 72 h. Six control animals were not sensitized but were injected with 0 1 ml of DNP-As (5 mg/ml), and the treated sites were observed for 72 h. The intensity of the inflammatory reactions was graded according to the method of Hayashi, Miyoshi, Nitta and Udaka
(1962).
Quantification of tissue leucocytosis At various intervals following challenge with antigen, the animals were killed, and the injection sites, previously marked, were immediately sampled using a punch (1 x 1 cm). The biopsy included all layers of the skin including the panniculus carnosus muscle. Eosinophil counts were made by the pre-
viously described method (Hirashima and Hayashi, 1976), using Luna's (1968) eosinophilic granule staining. From each 5-pm thick section, a total of fifteen random high power fields (10 x 40) were counted in 0-3-mm strips between the junction of the epidermis
and dermis and the upper limit of the panniculus carnosus muscle using a previously calibrated graticule. The cell count was expressed as a total of fifteen strips which represents the mean of the three sections counted, i.e. forty-five strips (Kay, 1970a). Macrophage experiments were performed essentially by the method of Turk, Heather and Diengdoh (1966). They demonstrated that macrophages could be differentiated from lymphocytes by the number and increased fragility of the lysosomes; and lysosomes were identified as cytoplasmic granules which stained for acid phosphatase. The skin tissue was placed with the fatty surface directly in contact with the inner wall of a glass vial in liquefied N2 gas. The vials were corked and stored at - 800 for not more than 3 days before cutting. Sections of each block were cut at 5 pm on a cryostat at - 300. Sections were attached to slides which were at room temperature and then fixed in pre-cooled 10 per cent neutral formalin at 40 for 2 h. After washing for 15 min, the sections were incubated at 370 with GomoriHolt acid phosphatase substrate mixture (Diengdoh and Turk, 1965) for 10, 20, 30 and 60 min, respectively. They were rinsed in distilled water and transferred to H2S-saturated water for 1 min. After washing, the sections were finally stained with haematoxylin and mounted in PVP. Macrophages were characterized by many granules which stained for acid phosphatase as described by Turk et al. (1966). On the other hand, such granules were scanty in lymphocytes. The quantification of such defined macrophages in the tissue was performed as described above. Estimation of chemotactic activity Before assay, all test samples were dialysed against 0 067 M phosphate buffer (pH 7 4) at 20 for 16 h. Chemotactic activity for eosinophils was measured in vitro by a modification (Yamamoto, Yoshinaga and Hayashi, 1971) of Boyden's method (1962) using Millipore filters (SCWP 29300, pore size 8 ,pm) (Millipore Filter Company, Bedford, Massachusetts) as previously described (Hirashima and Hayashi, 1976). The chambers containing two 1-ml compartments were incubated for 3 h at 370 using a 5 per cent CO2 atmosphere. The cells that had migrated through the filter on the lower surface of the filter were stained with Carazzi's haematoxylin and Chromotrope 2R (Kay, 1970b). The chemotactic counts were expressed as the mean count of migrated eosinophils on twenty high power fields (10 x 40)
265
Tissue eosinophilia in hypersensitivity. II.
randomly selected. Guinea-pig eosinophils were obtained by peritoneal lavage from animals which had received weekly multiple (eight to twelve) injections of 2 ml of horse serum (Litt, 1960), as previously described (Hirashima and Hayashi, 1976). The cells including 55-75 per cent eosinophils were suspended in Hanks's balanced salt solution containing 0-5 per cent ovalbumin at a concentration of 1 x 106 eosinophils/ml. Chemotactic assays for macrophages were made in vitro in the chambers as above, using Nuclepore filters (GE-500, pore size 5 jm) (General Electric Company, New York). The chambers were incubated at 370 for 90 min using a 5 per cent CO2 atmosphere (Horwitz and Garrett, 1971). The cells, which had migrated through the filter on the lower surface of the filter, were stained with Giemsa's reagent and counted on five high power fields (10 x 40) randomly selected. Guinea-pig macrophages were collected from the peritoneal cavity 4 days after instillation of mineral oil (Fishman, 1961) and suspended in RPMI 1640 solution, pH 7A4 (Grand Island Biological Company, Grand Island, New York) containing 10 per cent inactivated foetal calf serum (Microbiological Associates Incorporated, Bethesda, Maryland) at a concentration of 5 x 10s cells/ml. If necessary, migrated cells were confirmed to be macrophages by histochemical staining for acid phosphatase. Preparation of skin extract Extracts were made by the method of Hayashi et al. (1962), as previously described, immediately after the animals were killed. The skin (8-12 g per animal) was cut and frozen at - 80°. The pieces of frozen skin were cut into slices about 50 pm thick with a freezing microtome, the slices being dehydrated with three changes of cold acetone and powdered. Powdered skin (3-4 g per animal) was extracted with a ten-fold volume of 0-067 M phosphate buffer (pH 7A4) at 2° for 4 h. After centrifugation, the clear supernatant fluid was concentrated by ultrafiltration using Diaflo membrane, UM-2 (Amicon Company, Lexington) to give an absorbancy of 40 at 280 nm/ml.
Gelfiltration of skin extract on Sephadex G-100 The clear skin extract (5 ml, absorbancy 40 at 280 nm/ml) from 24-h-old lesions was applied to a column (3 x 50 cm) of Sephadex G-100, fine (Pharmacia, Uppsala, Sweden) equilibrated with 0-067 M
phosphate buffer (pH 7 4) (Porath and Flodin, 1959). The flow rate was 16 ml/h and 3-g effluent fractions were collected for assay of chemotactic activity. Gel filtration of chenotactic fraction on Sephadex G-200 The first component (stippled area in Fig. 3) (5 ml, absorbancy 20 at 280 nm/mi) after gel filtration on Sephadex G-100 was applied to a column (2-5 x 100 cm) of Sephadex G-200 (Pharmacia, Uppsala, Sweden) equilibrated with 0-067 M phosphate buffer (pH 7 4) (Andrews, 1965). The flow rate was 8 ml/h and 3-g effluent fractions were collected for assay of chemotactic activity. Furthermore, the second component (stippled area in Fig. 4) (5 ml, absorbancy 10 at 280 nm/ml) after gel filtration on Sephadex G-200 was re-chromatographed through the same column under the same conditions as described above. RESULTS Time-course of leucocyte reaction in active cutaneous anaphylaxis Intradermal injection of 0-1 ml of DNP-As (5 mg/ml) into sensitized animals induced inflammatory reactions, becoming maximal in about 18 h and persisting to some extent for 48-60 h, as previously described (Hirashima and Hayashi, 1976). In maximal reactions (graded + + to + + +), moderate
haemorrhage and sometimes mild tissue sloughing occurred in the lesion centres. The animals showing severe haemorrhage or tissue sloughing were not used in the present experiment. Tissue leucocytosis in the reaction was histologically and histochemically examined. As shown in Fig. la and b, macrophages could be differentiated from lymphocytes by many lysosomal cytoplasmic granules which stained for acid phosphatase; macrophages packed full of acid phosphatase containing granules in their cytoplasm, and lymphocytes were found to contain no or only a very few granules. Fibroblasts, although small in number, were occasionally found to contain granules but were easily distinguished by their shape. Macrophages were largely found around the small blood vessels, especially the venules and then infiltrated diffusely in varying degree. As illustrated in Fig. 2, infiltration of such defined macrophages in the
M. Hirashima, M. Honda & H. Hayashi
266
Figure 1. (a) Section from 24-h-old skin lesion induced by DNP-As in sensitized guinea-pig, stained for acid phosphatase. Macrophages are characterized by many lysosomal granules which stained for acid phosphatase. Such cells are largely found in the perivascular area. (Magnification x224.) (b) Higher magnification of macrophages full of acid phosphatase-positive granules in their cytoplasm. (Magnification x448.)
The mediation of tissue eosinophilia in hypersensitivity reactions II. SEPARATION OF A DELAYED EOSINOPHIL CHEMOTACTIC FACTOR FROM MACROPHAGE CHEMOTACTIC FACTORS
M. HIRASHIMA, M. HONDA & H. HAYASHI Department of Pathology, Kumamoto University Medical School, Kumamoto, Japan
Received 22 December 1975; accepted for publication 2 February 1976
Summary. In anaphylactic cutaneous lesions induced by DNP-ascaris extract in the guinea-pig, the timecourse of delayed tissue eosinophilia was found to parallel that of the macrophage reaction, reaching its peak in 24 h. Macrophages could be differentiated I from lymphocytes by the numerous lysosomal I granules which stained for acid phosphatase. Extracts from such skin lesions contained a delayedI eosinophil chemotactic factor and two different macrophage chemotactic factors. Most of the delayed eosinophil chemotactic factor was separated from the two macrophage chemotactic factors by gel filtration on Sephadex G-100 and Sephadex G-200 in that order. The eosinophil chemotactic factor after re-chromatography on Sephadex G-200 showed no or little chemotactic activity for macrophages. INTRODUCTION The isolation and characterization of chemotactic factors present in inflammatory tissues is essential in order to clarify the mediators of inflammatory leucotaxis, though there remains the difficult Correspondence: Professor H. Hayashi, Department of Pathology, Kumamoto University Medical School, Kumamoto 860, Japan.
263
problem of obtaining sufficient amounts of chemotactic factors from tissues. As previously described (Hirashima and Hayashi, 1976), in active cutaneous anaphylaxis induced by DNP-ascaris extract in guinea-pigs, tissue eosinophilia manifested two phases; the early phase seemed to be associated with a thermostable factor with a molecular weight of less than 1400, probably dependent of mast cell degranulation and resembling chemotactic factors described by Parish and Coombs (1968) and Kay, Stechschulte and Austen (1971), while the delayed phase with a thermolabile factor with a molecular weight of about 70,000, probably independent of mast cell degranulation and difficult to compare with previously proposed chemotactic factors (see review by Parish, 1974). The delayed factor was considered to be more significant than the early factor, because the delayed tissue eosinophilia was more intense and prolonged than the early tissue eosinophilia. Since the chemotactic assay in the previous experiment was performed with eosinophils only, it remained to be resolved whether the delayed factor was selective for eosinophils. The purpose of the present communication is to describe a mainly parallel rise in the numbers of eosinophils and macrophages in the delayed phase, and a possible separation of the delayed factor from the macrophage chemotactic factors in the same skin extract.
264
M. Hirashima, M. Honda & H. Hayashi
MATERIALS AND METHODS Animals Male albino guinea-pigs of Hartley strain weighing 400 ± 50 g were used for induction of active cutaneous anaphylaxis by DNP-ascaris extract. The animals weighing 600-1000 g were used for separation of eosinophils from the peritoneal exudate and those weighing 300 ± 20 g for separation of peritoneal macrophages. Induction of active cutaneous anaphylaxis Immunization of animals was done by a modification of the method of Margni and Hajos (1973). The animals were injected intramuscularly in both hind legs with 0-5 ml of DNP-ascaris extract (DNP-As) (5 mg/ml in physiological saline) prepared by the method of Eisen, Belman and Carsten (1953) and Tada and Okumura (1971), with an equal volume of Freund's complete adjuvant (Difco, Detroit, Michigan) and accompanied by simultaneous intraperitoneal injection of 1-0 ml of killed B. pertussis (1 x 10'°/ml in physiological saline). Seven days later, the same injections were repeated. Induction of active cutaneous anaphylaxis was performed following the previous method (Hirashima and Hayashi, 1976). Twelve days after the first immunizing injection, the twenty immunized animals from the experimental group were intracutaneously injected with 0-1 ml of DNP-As (5 mg/ml in physiological saline) at eight sites over the body, and the treated sites were observed for 72 h. Six control animals were not sensitized but were injected with 0 1 ml of DNP-As (5 mg/ml), and the treated sites were observed for 72 h. The intensity of the inflammatory reactions was graded according to the method of Hayashi, Miyoshi, Nitta and Udaka
(1962).
Quantification of tissue leucocytosis At various intervals following challenge with antigen, the animals were killed, and the injection sites, previously marked, were immediately sampled using a punch (1 x 1 cm). The biopsy included all layers of the skin including the panniculus carnosus muscle. Eosinophil counts were made by the pre-
viously described method (Hirashima and Hayashi, 1976), using Luna's (1968) eosinophilic granule staining. From each 5-pm thick section, a total of fifteen random high power fields (10 x 40) were counted in 0-3-mm strips between the junction of the epidermis
and dermis and the upper limit of the panniculus carnosus muscle using a previously calibrated graticule. The cell count was expressed as a total of fifteen strips which represents the mean of the three sections counted, i.e. forty-five strips (Kay, 1970a). Macrophage experiments were performed essentially by the method of Turk, Heather and Diengdoh (1966). They demonstrated that macrophages could be differentiated from lymphocytes by the number and increased fragility of the lysosomes; and lysosomes were identified as cytoplasmic granules which stained for acid phosphatase. The skin tissue was placed with the fatty surface directly in contact with the inner wall of a glass vial in liquefied N2 gas. The vials were corked and stored at - 800 for not more than 3 days before cutting. Sections of each block were cut at 5 pm on a cryostat at - 300. Sections were attached to slides which were at room temperature and then fixed in pre-cooled 10 per cent neutral formalin at 40 for 2 h. After washing for 15 min, the sections were incubated at 370 with GomoriHolt acid phosphatase substrate mixture (Diengdoh and Turk, 1965) for 10, 20, 30 and 60 min, respectively. They were rinsed in distilled water and transferred to H2S-saturated water for 1 min. After washing, the sections were finally stained with haematoxylin and mounted in PVP. Macrophages were characterized by many granules which stained for acid phosphatase as described by Turk et al. (1966). On the other hand, such granules were scanty in lymphocytes. The quantification of such defined macrophages in the tissue was performed as described above. Estimation of chemotactic activity Before assay, all test samples were dialysed against 0 067 M phosphate buffer (pH 7 4) at 20 for 16 h. Chemotactic activity for eosinophils was measured in vitro by a modification (Yamamoto, Yoshinaga and Hayashi, 1971) of Boyden's method (1962) using Millipore filters (SCWP 29300, pore size 8 ,pm) (Millipore Filter Company, Bedford, Massachusetts) as previously described (Hirashima and Hayashi, 1976). The chambers containing two 1-ml compartments were incubated for 3 h at 370 using a 5 per cent CO2 atmosphere. The cells that had migrated through the filter on the lower surface of the filter were stained with Carazzi's haematoxylin and Chromotrope 2R (Kay, 1970b). The chemotactic counts were expressed as the mean count of migrated eosinophils on twenty high power fields (10 x 40)
265
Tissue eosinophilia in hypersensitivity. II.
randomly selected. Guinea-pig eosinophils were obtained by peritoneal lavage from animals which had received weekly multiple (eight to twelve) injections of 2 ml of horse serum (Litt, 1960), as previously described (Hirashima and Hayashi, 1976). The cells including 55-75 per cent eosinophils were suspended in Hanks's balanced salt solution containing 0-5 per cent ovalbumin at a concentration of 1 x 106 eosinophils/ml. Chemotactic assays for macrophages were made in vitro in the chambers as above, using Nuclepore filters (GE-500, pore size 5 jm) (General Electric Company, New York). The chambers were incubated at 370 for 90 min using a 5 per cent CO2 atmosphere (Horwitz and Garrett, 1971). The cells, which had migrated through the filter on the lower surface of the filter, were stained with Giemsa's reagent and counted on five high power fields (10 x 40) randomly selected. Guinea-pig macrophages were collected from the peritoneal cavity 4 days after instillation of mineral oil (Fishman, 1961) and suspended in RPMI 1640 solution, pH 7A4 (Grand Island Biological Company, Grand Island, New York) containing 10 per cent inactivated foetal calf serum (Microbiological Associates Incorporated, Bethesda, Maryland) at a concentration of 5 x 10s cells/ml. If necessary, migrated cells were confirmed to be macrophages by histochemical staining for acid phosphatase. Preparation of skin extract Extracts were made by the method of Hayashi et al. (1962), as previously described, immediately after the animals were killed. The skin (8-12 g per animal) was cut and frozen at - 80°. The pieces of frozen skin were cut into slices about 50 pm thick with a freezing microtome, the slices being dehydrated with three changes of cold acetone and powdered. Powdered skin (3-4 g per animal) was extracted with a ten-fold volume of 0-067 M phosphate buffer (pH 7A4) at 2° for 4 h. After centrifugation, the clear supernatant fluid was concentrated by ultrafiltration using Diaflo membrane, UM-2 (Amicon Company, Lexington) to give an absorbancy of 40 at 280 nm/ml.
Gelfiltration of skin extract on Sephadex G-100 The clear skin extract (5 ml, absorbancy 40 at 280 nm/ml) from 24-h-old lesions was applied to a column (3 x 50 cm) of Sephadex G-100, fine (Pharmacia, Uppsala, Sweden) equilibrated with 0-067 M
phosphate buffer (pH 7 4) (Porath and Flodin, 1959). The flow rate was 16 ml/h and 3-g effluent fractions were collected for assay of chemotactic activity. Gel filtration of chenotactic fraction on Sephadex G-200 The first component (stippled area in Fig. 3) (5 ml, absorbancy 20 at 280 nm/mi) after gel filtration on Sephadex G-100 was applied to a column (2-5 x 100 cm) of Sephadex G-200 (Pharmacia, Uppsala, Sweden) equilibrated with 0-067 M phosphate buffer (pH 7 4) (Andrews, 1965). The flow rate was 8 ml/h and 3-g effluent fractions were collected for assay of chemotactic activity. Furthermore, the second component (stippled area in Fig. 4) (5 ml, absorbancy 10 at 280 nm/ml) after gel filtration on Sephadex G-200 was re-chromatographed through the same column under the same conditions as described above. RESULTS Time-course of leucocyte reaction in active cutaneous anaphylaxis Intradermal injection of 0-1 ml of DNP-As (5 mg/ml) into sensitized animals induced inflammatory reactions, becoming maximal in about 18 h and persisting to some extent for 48-60 h, as previously described (Hirashima and Hayashi, 1976). In maximal reactions (graded + + to + + +), moderate
haemorrhage and sometimes mild tissue sloughing occurred in the lesion centres. The animals showing severe haemorrhage or tissue sloughing were not used in the present experiment. Tissue leucocytosis in the reaction was histologically and histochemically examined. As shown in Fig. la and b, macrophages could be differentiated from lymphocytes by many lysosomal cytoplasmic granules which stained for acid phosphatase; macrophages packed full of acid phosphatase containing granules in their cytoplasm, and lymphocytes were found to contain no or only a very few granules. Fibroblasts, although small in number, were occasionally found to contain granules but were easily distinguished by their shape. Macrophages were largely found around the small blood vessels, especially the venules and then infiltrated diffusely in varying degree. As illustrated in Fig. 2, infiltration of such defined macrophages in the
M. Hirashima, M. Honda & H. Hayashi
266
Figure 1. (a) Section from 24-h-old skin lesion induced by DNP-As in sensitized guinea-pig, stained for acid phosphatase. Macrophages are characterized by many lysosomal granules which stained for acid phosphatase. Such cells are largely found in the perivascular area. (Magnification x224.) (b) Higher magnification of macrophages full of acid phosphatase-positive granules in their cytoplasm. (Magnification x448.)
