Granulocyte-Macrophage Colony-Stimulating Factor Produced by ...

Vol. 51, No. 1

INFECTION AND IMMUNITY, Jan. 1986, p. 213-217

0019-9567/86/010213-05$02.00/0 Copyright © 1986, American Society for Microbiology

Granulocyte-Macrophage Colony-Stimulating Factor Produced by Splenic T Lymphocytes of Mice Infected with Schistosoma japonicum MAKOTO OWHASHI AND YUKIFUMI NAWA*

Department of Parasitology, Miyazaki Medical College, Kiyotake, Miyazaki 889-16, Japan Received 5 September 1985/Accepted 8 October 1985

The production of granulocyte-macrophage (GM) colony-stimulating factor (CSF) by splenic lymphocytes examined in murine schistosomiasis japonica. When splenic lymphocytes obtained at various weeks after infection were cultured with soluble egg antigen, GM-CSF activity in the conditioned medium became detectable at 3 weeks after infection, reached a peak at week 5, and persisted at least up to week 7. Not only soluble egg antigen but also concanavalin A was highly effective in stimulating splenic lymphocytes to produce GM-CSF. When splenic lymphocytes were treated with anti-Thy-1.2 antibody and complement, GM-CSFproducing activity was completely abolished. The molecular weight of this T-cell-derived GM-CSF was estimated to be 30,000 by gel ifitration on Sephadex G-150. After isoelectric focusing, GM-CSF activity was detected as two major peaks at pH 3.7 and 5.5. The physicochemical nature of this T-cell-derived GM-CSF was compared with those of known lymphokine GM-CSFs or with that of a previously reported GM-CSF in the serum of S. japonicum-infected mice (M. Owhashi and Y. Nawa, Infect. Immun. 49:533-537, 1985).

was

Co.). After being washed with saline, the eggs were suspended in phosphate-buffered saline and homogenized with a Teflon homogenizer. After overnight extraction at 4°C, the mixture was centrifuged at 100,000 x g for 1 h. The supernatant was used as crude SEA. Preparation of conditioned medium. The spleens were removed from groups of three to five mice 6 weeks after infection, except for the kinetic study. They were gently squashed between two frost-ended slides in cold Hanks balanced salt solution. The cell suspensions were washed with Hanks balanced salt solution and suspended in RPMI 1640 (GIBCO Laboratories) supplemented with 2% heatinactivated fetal bovine serum (Flow Laboratories, Inc.) and

Granulocyte-macrophage (GM) colony-stimulating factor (CSF) not only is a myelopoietin required for the proliferation and differentiation of normal GM precursor cells in vitro (1) but also has a larger role in regulating the macrophagemononuclear phagocyte system (8). As sources of GM-CSF, monocytes-macrophages (4) or fibroblasts (16) are well known. In addition, T lymphocytes are, after stimulation with various mitogens (13) or a mixed lymphocyte reaction (14), capable of producing GM-CSF. Furthermore, GM-CSF production by a T-cell hybridoma (2) or thymoma cell line EL-4 (6) after stimulation with concanavalin A (ConA) has been reported. Recently, we found that a Schistosoma japonicum infection in mice caused a rapid and relatively persistent rise in GM-CSF levels in serum (12). Since splenomegaly is one of the characteristic features of schistosomiasis (17) and since the splenic lymphocytes of Schistosoma-infected mice have been shown to produce certain kinds of lymphokines, such as eosinophil chemotactic factor (7), we tested the GM-CSFproducing activity of the splenic lymphocytes of S. japonicum-infected mice. The results showed that splenic T cells are, after stimulation with soluble egg antigen (SEA), able to produce GM-CSF. The physicochemical nature of this GM-CSF was compared with those of known T-cellderived GM-CSFs or with that of a previously reported serum GM-CSF (M. Owhashi and Y. Nawa, Infect. Immun. 49:533-537, 1985).

5

MATERIALS AND METHODS Mice and infection. Male C57BL/6 mice, 5 to 6 weeks old and weighing approximately 17 g, were infected with S. japonicum (Kofu strain) by intraperitoneal injection with 30 cercariae.

Preparation of SEA. SEA was prepared by methods described previously (11). In short, eggs were harvested from the intestines of infected mice by enzymatic digestion with pronase (Kaken) and collagenase (type I; Sigma Chemical *

x

10-5 M 2-mercaptoethanol. For removal of adherent

cells, the cell suspensions were plated in plastic dishes (Falcon 3001; Becton Dickinson Labware) and incubated at 37°C for 1 h in a 5% C02-air environment. Nonadherent cells were harvested, washed, and suspended in fresh medium at a cell concentration of 2 x 106/ml, except as otherwise stated. Culturing was carried out at 37°C for 24 h in a 5% C02-air environment. Anti-Thy-1.2 treatment was performed as follows. Plastic-nonadherent splenic lymphocytes were washed and suspended in RPMI 1640 containing 0.3% bovine serum albumin and 25 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) (Cederlane Cytotoxicity Medium) and various concentrations of antiThy-1.2 monoclonal antibody (F7D5 immunoglobulin M cytotoxic monoclonal antibody; Serotec). After incubation at room temperature for 30 min, the cells were washed with cytotoxicity medium and resuspended in the same medium containing appropriately diluted complement (Low-Tox-M Rabbit Complement; Cederlane). After incubation at 37°C for 40 min, the cells were washed and suspended in complete culture medium. Conditioned medium was obtained by centrifugation at 1,200 x g for 10 min, sterilized by filtration through a 0.45-,um membrane filter (Millipore Corp.), and stored at -30°C until used.

Corresponding author. 213

OWHASHI AND NAWA

214

INFECT. IMMUN.

0.3

1.0

3.0

10 0 10 10i 10' jug/ml celIsImI FIG. 1. Dose-resionse relationship between GM-CSF activity and the concentration of SEA added or the number of cells in the culture. Spleen cells were obtained 6 weeks after infection with 30 cercariae of S. japonicum. (A) A fixed number (2 x 106/ml) of splenic lymphocytes was cultured with various concentrations of SEA (0). Control cultures contained various concentrations of SEA alone (0). (B) Various numbers of splenic lymphocytes were cultured with 5 ,ug of SEA per ml (0) or without SEA (0). Vertical bars represent the standard 0

deviation of the mean.

In vitro colony assay. Details of the techniques for colony formation in soft agar have been described previously (12). Gel filtration. Sephadex G-150 (Pharmacia, Inc.) was prepared as a column with bed dimensions of 1.6 by 65 cm and was equilibrated with 5.8 mM phosphate-buffered saline (pH 7.4) containing 0.02% Tween 20 (10). Elution was carried out with the same buffer at a flow rate of 4 ml/h at 4°C, and 2.2-ml fractions were collected. Blue dextran (Pharmacia), aldolase (Sigma), bovine serum albumin (Sigma), ovalbumin (Sigma), a-chymotrypsinogen (Sigma), and cytochrome c (Sigma) were used as molecular weight markers. Isoelectric focusing. Isoelectric focusing was carried out with a 110-ml column by a method described previously (11). The pH gradient was formed with ampholites (pH 3.5 to 10; LKB Productor). After focusing at 0C, 1.9-ml fractions were collected, and the pH was measured in an ice-chilled water bath. Each fraction was dialyzed against phosphatebuffered saline before examination of CSF activity. RESULTS Dose response of cells and SEA. To determine the optimal conditions for CSF production, 2 x 106 splenic lymphocytes

50

obtained 6 weeks after S. japonicum infection were cultured with various concentrations of SEA (Fig. 1A) or, alternatively, various numbers of cells were cultured with 5 p.g of SEA per ml (Fig. 1B). When fixed number of cells were cultured with 0 to 10 ,ug of SEA per ml, CSF activity in the culture supernatant increased almost linearly with the dose of SEA and reached a plateau at 3 jig of SEA per ml (Fig. 1A). When various numbers of splenic lymphocytes were cultured with 5 pg of SEA per ml, optimal CSF activity was observed at around 3 x 106 cells per ml (Fig. 1B). CSF activity was not detected in the supernatants of the negative control cultures of SEA alone (Fig. 1A) or cells alone (Fig. 1B). Time course study. Splenic lymphocytes (2 x 106/ml) obtained from individual mice at various times after infection with S. japonicum were cultured with or without 5 ,ug of SEA per ml. The conditioned medium was harvested 24 h later, and their CSF activity was measured (Fig. 2). When splenic lymphocytes were cultured with SEA, CSF activity became detectable at 3 weeks after infection, reached a peak at week 5, and persisted at least up to week 7. Morphologically, the colonies generated in these experiments were of GM lineage (data not shown). Stimulation of CSF production by ConA. Splenic lymphocytes from normal or S. japonicum-infected mice (6 weeks after infection) were stimulated with either 3 jig of a specific antigen (SEA) or 1 ,ug of ConA per ml, and the CSF activity

.U 0 U)

0

Number of colonies/dish 50 100

0

o40 25

Con A Normal

SEA

E z

Con A 0

2

4

6 a Weeks post-infection FIG. 2. Kinetic changes in the GM-CSF-producing activity of splenic lymphocytes after infection. Splenic lymphocytes obtained from individual mice at various times after infection were cultured with (0) or without (0) 5 pg of SEA per ml. Each point represents the mean for three mice. Vertical bars represent the standard deviation of the mean.

Infec ted

SEA

I ±-4

FIG. 3. Effects of ConA on the GM-CSF-producing activity of splenic lymphocytes. Splenic lymphocytes (2 x 106/ml) from normal or S. japonicum-infected mice (6 weeks postinfection) were cultured with 3 ,ug of SEA or 1 ,ug of ConA per ml. Each column represents the mean + standard deviation (vertical bars) of triplicate assays.

LYMPHOKINE GM-CSF IN MURINE SCHISTOSOMIASIS

VOL. 51, 1986

.C 100 U)

0

n 50 E

z

0

(~)

ft

*.

2x10-5

2x10-6

2x10-4

2x10-3

215

,ug of SEA per ml, and the conditioned medium was concentrated and applied to a Sephadex G-150 column. The elution pattern is shown in Fig. 5. CSF activity was detected at the elution position between ovalbumin and a-chymotrypsinogen. From the calibration curve (Fig. 6A), the apparent molecular weight of the lymphocyte-derived CSF was estimated to be 30,000. As a reference, a calibration curve prepared from our previously reported results of GM-CSF in the serum of S. japonicum-infected mice (12) is also shown (Fig. 6B). After isoelectric focusing, the lymphocyte-derived CSF was detected as two major peaks at pH 3.7 and 5.5, with several minor components (Fig. 7A). On the other hand, GM-CSF in the serum of S. japonicum-infected mice had isoelectric points of 4.0 and 7.0 (Fig. 7B).

Anti-Thy-i dilution

FIG. 4. Effects of anti-Thy-1.2 treatment on CSF-producing activity. Splenic lymphocytes were treated with various concentrations of anti-Thy-1.2 monoclonal antibody and complement. The cells were then stimulated with 5 ,ug of SEA per ml, and the CSF activity in the conditioned medium was examined.

in the conditioned medium was examined (Fig. 3). Both SEA and ConA stimulated splenic lymphocytes from infected mice to produce CSF. On the other hand, only ConA stimulated normal splenic lymphocytes. Effects of anti-Thy-1.2 treatment on CSF-producing activity. Since splenic lymphocytes from S. japonicum-infected mice produced CSF after stimulation with a specific antigen or ConA, T lymphocytes seemed to be the most probable source of CSF. To confirm this possibility, we examined the effects of anti-Thy-1.2 treatment on the CSF-producing activity of splenic lymphocytes from infected mice (Fig. 4). When 2 x 106 splenic lymphocytes per ml were treated with anti-Thy-1.2 antibody and complement and then stimulated with 5 ,ug of SEA per ml, the complete inhibition of CSF production was observed at 20,000-fold or higher concentrations of antibody. Physicochemical properties of GM-CSF. Splenic lymphocytes (2 x 106/ml) from infected mice were cultured with 5

DISCUSSION The results reported here show that GM-CSF is detectable in the conditioned medium prepared by culturing spleen cells obtained from S. japonicum-infected mice. CSF production is antigen dependent but is also stimulated by ConA. Furthermore, CSF-producing activity was completely abolished by anti-Thy-1.2 treatment, indicating that this CSF is a T-cell-derived lymphokine. In addition, the apparent molecular weight and isoelectric points were similar to those of previously reported GM-CSFs derived from a T-cell line (6) and a T-cell hybridoma (2). The kinetics of GM-CSF-producing activity of spleen cells after S. japonicum infection were essentially similar to those of the immune and mitogen-stimulated blastogenic responses of S. japonicum-infected mouse spleen cells reported by Garb et al. (5) but were somewhat different from those of the mitogen-stimulated blastogenic responses reported by Warren et al. (18). This may be due to the degree of infection, because we used 30 cercariae per mouse, an amount similar to that used by Garb et al. (25 cercariae per mouse) but different from that used by Warren et al. (5 cercariae per mouse). Since we recently reported a rise in GM-CSF levels in the serum of S. japonicum-infected mice (12), the question arises as to whether the GM-CSF in serum and that pro-

I

.1

-C

Ec

U)

U)

0

.

0 Go V C '0

7D 0

E

.0

40

z

80 Fraction number FIG. 5. Sephadex G-150 gel chromatography. Splenic lymphocytes (2 x 106/ml) obtained 6 weeks after infection were cultured with 5 ,ug of SEA per ml. The conditioned medium was concentrated 30 times with a Diaflow cell equipped with a UM-10 membrane. Two milliliters of the concentrated material was applied to the Sephadex G-150 column. GM-CSF activity is expressed as the number of colonies generated per dish. Blue dextran (BD), aldolase (Ald), bovine serum albumin (BSA), ovalbumin (OA), a-chymotrypsinogen (Chm), cytochrome c (Cyt), and NaCl were eluted at the position indicated at the top of the figure.

20

40

60

INFECT. IMMUN.

OWHASHI AND NAWA

216

IA

ioe6B Fer

Aid .0.

BSA QA CsF

co

CSFa

Chi

BSA OA

yt

Z 104

1.0

0.5

0

Kav

0

0.5

1.0 Kav

FIG. 6. Molecular weight calibration. (A) Calibration curve for estimation of the molecular weight of GM-CSF derived from splenic lymphocytes. The curve was prepared from the results of Sephadex G-150 gel chromatography (Fig. 5). The apparent molecular weight of lymphocyte-derived GM-CSF was estimated to be 30,000. (B) Calibration curve for estimation of the molecular weight of GM-CSF in the serum of S. japonicum-infected mice. The curve was prepared from the results of Sephadex G-200 gel chromatography (12). The apparent molecular weight of serum-derived GM-CSF was estimated to be 260,000. Abbreviations for the molecular weight markers are the same as those in Fig. 5. In addition, ferritin (Fer) and catalase (Cat) were used for the calibration of the Sephadex G-200 column.

duced by sensitized T cells in murine schistosomiasis are identical or not. When kinetic changes in serum CSF levels and in the ability of T cells to produce CSF are compared, they are roughly the same. However, the apparent molecular weights of these two CSFs are markedly different. Thus, one possible explanation is that T-cell-derived CSF is bound to a carrier molecule in the circulation. Related to this, several workers (Sa, 9) reported that apparent-high-molecularweight CSF in serum could be reduced by neuraminidase treatment and gel filtration under dissociating conditions. Alternatively, CSF in serum may be derived from cells other than T cells. If this is the case, macrophages are the most likely source of high-molecular-weight CSF in serum. The rise in serum CSF is caused not only by various infections

A

Of

10

100 8 6 4

0

2: 10

20

30

40

_

CA

I0 5i 30 40 Fraction number FIG. 7. Isoelectric focusing of GM-CSF derived from splenic lymphocytes (A) and from the serum of S. japonicum-infected mice (B). Lymphokine-derived CSF had two major peaks (pH 3.7 and 5.5) and several minor peaks. Serum-derived CSF also had two major peaks (pH 4.0 and 7.0). 10

20

but also by stimulation with bacterial endotoxin (3) or by estrogen (5a), both of which are well-known stimulators of the mononuclear phagocyte system (15). Further physicochemical characterization and knowledge of the functional role, other than as myelopoietin, of these two GM-CSFs would elucidate the relationship between them. ACKNOWLEDGMENTS We are grateful to T. Hayama (Department of Anatomy, Kumamoto University Medical School) for his valuable advice and constructive criticism. We thank Eri Tandou for her excellent technical assistance. This work was supported by a Grant-in-Aid for Special Project Research from the Ministry of Education, Science and Culture, Japan. LITERATURE CITED 1. Bradley, T. R., and D. Metcalf. 1966. The growth of mouse bone marrow cells in vitro. Aust. J. Exp. Biol. Med. Sci. 44:287-300. 2. Burgess, A. W., P. F. Bartlett, D. Metcalf, N. A. Nicola, I. Clark-Lewis, and J. W. Schrader. 1981. Granulocyte-macrophage colony-stimulating factor produced by an inducible murine T-cell hybridoma: molecular properties and cellular specificity. Exp. Hematol. (N.Y.) 9:893-903. 3. Chervenick, P. A. 1972. Effect of endotoxin and postendotoxin plasma on in vitro granulopoiesis. J. Lab. Clin. Med. 79: 1014-1020. 4. Eaves, A. C., and W. R. Bruce. 1974. In vitro production of colony-stimulating activity. I. Exposure of mouse peritoneal cells to endotoxin. Cell Tissue Kinet. 7:19-30. 5. Garb, K. H., A. B. Stavitsky, and A. A. F. Mahmoud. 1981. Dynamics of antigen and mitogen-induced responses in murine schistosomiasis japonica: in vitro comparison between hepatic granulomas and splenic cells. J. Immunol. 127:115-120. 5a.Hayama, T., Y. Nawa, and M. Kotani. 1985. Granulocytemacrophage colony stimulating activity (GM-CSA) in the serum of estriol-treated mice. Exp. Hematol. 13:658-663. 6. Hilfiker, M. L., R. N. Moore, and J. J. Farrar. 1981. Biologic properties of chromatographically separated murine thymomaderived interleukin 2 and colony-6timulating factor. J. Immunol. 127:1983-1987. 7. Lewis, F. A., C. E. Carter, and D. G. Colley. 1977. Eosinophils and immune mechanisms. V. Demonstration of mouse spleen cell-derived chemotactic activities for eosinophils and mononuclear cells and comparisons with eosinophil stimulating promoter. Cell. Immunol. 32:8696. 8. Moore, R. N., J. T. Hoffeld, J. J. Farrar, S. E. Mergenhagen,

VOL. 51, 1986

9.

10.

11.

12. 13.

LYMPHOKINE GM-CSF IN MURINE SCHISTOSOMIASIS

J. J. Oppenhein, and R. K. Shadduck. 1981. Role of colonystimulating factors as primary regulators of macrophage functions. Lymphokines 3:119-148. Nicola, N. A., A. W. Burgess, and D. Metcalf. 1979. Similar molecular properties of granulocyte-macrophage colony-stimulating factors produced by different mouse organs in vitro and in vivo. J. Biol. Chem. 254:5290-5299. Nicola, N. A., D. Metealf, M. Matsumoto, and G. R. Johnson. 1983. Purification of a factor inducing differentiation in murine myelomonocytic leukemia cells. Identification as granulocyte colony-stimulating factor. J. Biol. Chem. 258:9017-9023. Owhashi, M., and A. Ishii. 1982. Purification and characterization of a high molecular weight eosinophil chemotactic factor from Schistosomajaponicum eggs. J. Immunol. 129:2226-2231. Owhashi, M., and Y. Nawa. 1985. Granulocyte-macrophage colony-stimulating factor in the sera of Schistosomajaponicuminfected mice. Infect. Immun. 49:533-537. Parker, J. W., and D. Metcalf. 1974. Production of colony-

14.

15.

16. 17.

18.

217

stimulating factor in mitogen-stimulated lymphocyte cultures. J. Immunol. 112:502-510. Parker, J. W., and D. Metcalf. 1974. Production of colonystimulating factor in mixed leucocyte cultures. Immunology 26:1039-1049. Stuart, A. E., J. A. Habeshaw, and A. E. Davidson. 1978. Phagocytes in vitro, p. 31.1-31.30. In D. M. Weir (ed.), Handbook of experimental immunology, 3rd ed. Blackwell Scientific Publications, Ltd., Oxford. Waheed, A., and R. K. Shadduck. 1979. Purification and properties of L cellderived colony stimulating factor. J. Lab. Clin. Med. 94:180-194. Warren, K. S. 1973. The pathology of schistosome infections. Helminthol. Abstr. 42:591-633. Warren, K. S., D. I. Grove, and R. P. Pelly. 1978. The Schistosoma japonicum egg granuloma. II. Cellular composition, granuloma size, and immunologic concomitants. Am. J. Trop. Med. Hyg. 27:271-275.