- Email: [email protected]
Synthesis of factor D by gastric cancer-derived cell lines
International Immunopharmacology 2 (2002) 843 – 848 www.elsevier.com/locate/intimp
Synthesis of factor D by gastric cancer-derived cell lines Etsuko Kitano, Hajime Kitamura* Department of Clinical Laboratory Science, Osaka Prefecture College of Health Sciences, Habikino, Osaka, Japan Received 18 June 2001; received in revised form 25 February 2002; accepted 11 March 2002
Abstract Synthesis of complement components in vitro by four human gastric cancer-derived cell lines, MKN28, MKN74, MKN45 and KATO-III, was studied. When these cells were cultured for 3 days without addition of any stimulator, 0.94F0.49, 2.10F0.59, 7.29F5.94 and 2.47F1.34 ng of factor D/106 cells were detected in supernatants of MKN28, MKN74, MKN45 and KATO-III, respectively. Factor D production by these cells was reversibly inhibited by the presence of cycloheximide. Factors B, C3 and C2 were also detected in protein-free culture medium of these cell lines. Addition of tumour necrosis factor (TNF) to culture enhanced C3 and factor B secretion but depressed C2 secretion, without any distinct effect on factor D secretion. Since all cell lines tested secreted significant amounts of factor D without addition of any stimulator in medium, it is possible that factor D may be synthesized by gastric epithelial cells physiologically and constitutively. From a quantitative analysis of factor D secretion by these cells, factor D secreted by gastric tissue is likely to contribute to the factor D level in circulating blood. The possible mechanism of participation of complement system in inflammation of gastric epithelium was proposed. Thus, the present study may be significant for clarification of the mode of extrahepatic complement synthesis participating in mucosal immunity. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Complement; Factor D; Gastric epithelial cell; TNF
1. Introduction A number of studies related to the synthesis of complement components revealed that various kinds of cells other than hepatocytes can also produce complement components [1]. It is now believed that extrahepatic production of complement may contribute to complement levels in plasma and that local
*
Corresponding author. Tel.: +81-729-50-2111x3031; fax: +81-729-50-2126. E-mail address: [email protected] (H. Kitamura).
production of complement is important in homeostasis and immune defense in tissue [2]. We reported that all four kinds of cell lines derived from human gastric cancer tissue synthesize C3 in response to TNF, suggesting physiological or pathological production of C3 by gastric tissue [3]. This finding supports the hypothesis of Naughton et al. [4] that a significant proportion of the total systemic pool of C3 is synthesized extrahepatically and that the epithelium of the gastrointestinal tract is likely to be responsible for C3 of extrahepatic origin. The present study deals with the secretion of factor D as well as factors B, C2 and C3 by human gastric cancer-derived cell lines.
1567-5769/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 7 - 5 7 6 9 ( 0 2 ) 0 0 0 2 8 - 0
844
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
2. Materials and methods 2.1. Reagents and antibodies Reagents and antibodies were obtained as follows: RPMI 1640, MEM and protein-free medium (PFHMII) were from Gibco (Grand Island, NY); fetal calf serum (FCS) was from CC Laboratories (Cleveland, OH); natural human TNF-a was from Hayashibara (Okayama, Japan); IL-1h and IL-6 were from Genzyme (Boston, MA); IFN-g was from Shionogi (Osaka, Japan); LPS was from Paesel (Germany). Antibody to human C3 (goat anti-C3 IgG) was from Cappel (West Chester, PA); rabbit antiserum to human C3c (rabbit anti-C3c) and to human C2 (rabbit anti-C2) were from Nordic, The Netherlands; sheep anti-factor D and sheep anti-factor B were from Binding Site (Birmingham, UK); rabbit anti-factor D was from Wako (Osaka); rabbit anti-factor B was from DAKO (Denmark); horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG and HRPconjugated rabbit anti-goat IgG and HRP-conjugated goat anti-rabbit IgG were from Bio-Rad (Richmond, CA). Sera from 40 healthy donors were combined and used as pooled normal human serum (NHS). Isolated factor D was kindly provided by Dr. K.S. Hong (JCR Pharmaceuticals, Kobe, Japan) and isolated factor B was purchased from Sigma (St. Louis, MO). C2 [5] and C3 [6] were isolated in our laboratory as described earlier. 2.2. Cell lines Four cell lines derived from human gastric cancer, MKN28, MKN45, MKN74 and KATO-III, were provided by The Japanese Cancer Research Resources Bank [3]. The histological origin of KATO-III cells are signet-ring carcinoma [7] and that of the other cell lines are adenocarcinoma [3]. Cells were grown in medium containing RPMI 1640 (45% v/v), MEM (45% v/v) and FCS (10% v/v).
plates (Falcon). These cell cultures were maintained in a humidified atmosphere of 5% CO2/95% air at 37 jC for 1 – 5 days. The culture medium was then harvested, centrifuged to remove cellular debris and maintained at 70 jC prior to study. 2.4. Complement assay ELISA for factors D and C3: The factor D protein [8] and the C3 protein [3] were quantitated by the sandwich ELISA technique. The factor B protein was also quantitated by the sandwich ELISA by using sheep anti-factor B, rabbit anti-factor B and HRPconjugated anti-rabbit IgG. Functional assay for C2: C2 activity was assayed by immune hemolysis as reported previously [9]. 2.5. Immunoblot analysis The culture supernatants, together with isolated component and NHS, were subjected to 2 – 15% SDS-PAGE under non-reducing conditions and transferred onto nitrocellulose according to the method of Towbin et al. [10]. After blocking unreacted sites and washing, blots were developed using the corresponding antibody and HRP-conjugated anti-IgG and the HRP substrate (Konica Immunostaining HRP1000). 2.6. Statistical analysis All data of concentrations of complement components in culture supernatants are expressed as the meanFS.D. of at least two experiments in triplicates. Data were analysed for statistical significance using Friedman test (StatView 5.0). Differences were considered significant for p<0.001.
3. Results 3.1. Factor D secretion by human gastric cancerderived cancer cell lines
2.3. Cell culture conditions To monitor complement secretion, cells were washed and cultured in fresh protein-free medium, with or without cytokine, in sterile 12-well culture
Gastric cancer-derived cell lines were cultured in PFHM-II without addition of any stimulator. Culture supernatants were sampled and assessed for antigenic factor D by ELISA. Fig. 1 clearly shows that these
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
845
trations of TNF-a for 2 days. Culture supernatants were assayed for factor B, factors D and C3 by ELISA and for hemolytic activity of C2. Results are shown in Fig. 3. Statistical analysis by Friedman test clearly showed that TNF-a enhanced factor B secretion and C3 secretion but suppressed C2 secretion by all of these cell lines in a dose-dependent fashion, while TNF-a did not show any distinct effect on factor D secretion. 3.5. Immunoblot analysis of factor D and factor B in culture supernatants
Fig. 1. Factor D secretion by human gastric cancer-derived cell lines. MKN28 cells (D-D), MKN74 cells (E-E), MKN45 cells (.-.) or KATO-III cells (5-5) were cultured in PFHM-II without addition of any cytokine. Culture supernatants were collected after culture for 1 – 5 days and assayed for factor D. Data represents meanFS.D. of two experiments in triplicates.
Culture supernatants of MKN45, MKN74 and KATO-III were applied for immunoblot analysis of factor D and factor B. As shown in Fig. 4, factor D secreted by MKN45, MKN74 or KATO-III exhibited a single band corresponding to a molecular weight similar to factor D in NHS and isolated factor D. Factor B secreted by these gastric cancer-derived cell lines also exhibited a single band of a molecular
cells produced increasing amounts of factor D over a 5-day period ( p<0.001 by Friedman test). 3.2. Effect of cycloheximide on factor D production Factor D secretion by MKN28 and KATO-III cell lines were completely blocked by cycloheximide treatment and was restored after removal of the inhibitor (Fig. 2). Similar results were obtained with other cell lines (not shown). These results indicate that factor D is synthesized de novo. 3.3. Effect of IFN-g, IL-1h and IL-6 on factor D secretion Factor D was assayed in culture supernatants obtained after incubation for 2 days in PFHM-II containing varying amounts of IL-6 (0 –300 ng/ml), IL-1h (0 – 33 ng/ml) or IFN-g (0 – 50,000 u/ml). Neither of these cytokines showed any distinct effect on factor D production (not shown). 3.4. Effect of TNF-a on factor B, factor D, C3 and C2 secretion MKN28 cells, MKN45 cells and KATO-III cells were cultured in PFHM-II containing varying concen-
Fig. 2. Effect of cycloheximide on factor D secretion. MKN28 cells or KATO-III cells were incubated in the presence (n) or absence (n) of cycloheximide (CHX, 1.0 Ag/ml) for 2 days, after which the cells were washed and incubated for another 2 days in the medium without CHX. Culture supernatants were collected and assayed for factor D. Data represents meanFS.D. of two experiments in triplicates.
846
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
Fig. 4. Immunoblot analysis of factor D in culture supernatants. Culture supernatants of KATO-III cells (K), MKN28 cells (28) and MKN74 cells (74), together with isolated factor D and NHS, were electrophoresed on a 2 – 15% SDS-PAGE and blotted onto nitrocellulose paper. The blot was analyzed for factor D.
weight similar to factor B in NHS and isolated factor B (not shown). These observations suggest that factors B and D molecules secreted by these cell lines might be similar to factors B and D in NHS.
4. Discussion
Fig. 3. Effect of TNF-a on the secretion of C3, C2, factor B and factor D. MKN28 cells (D-D), MKN45 cells (.-.) or KATO-III cells (5-5) were cultured in PFHM-II containing varying concentrations of TNF-a for 2 days. Culture supernatants were collected and assayed for factor B, factors D and C3 by ELISA and for hemolytic activity of C2. Data represents meanFS.D. of two experiments in triplicates.
Factor D has been shown to be identical to adipsin, which is synthesized almost exclusively by adipocytes [1,11,12]. However, human monocytes/macrophages [13], HepG2 cells [14] and astrocytes [15] were also reported to synthesize factor D. Recently, we reported that normal human hepatocytes secrete a significant amount of factor D in culture supernatants, suggesting that liver might be a major source of factor D in serum [8]. The present study clearly shows that factor D is secreted by gastric cancer-derived cell lines. This may be the first report of factor D synthesis by gastric epithelial cells. Of interest is that all of the tested cell lines secreted factor D without addition of any stimulator, suggesting the possibility that gastric epithelial cells might produce factor D constitutively. As to the amount of secreted factor D in culture supernatants, we have found that 106 gastric cells produced factor D at the rate of up to 1.0 ng/day (Fig. 1), corresponding to 0.1% of the protein present in 1.0 ml of serum (1.0 –2.0 Ag/ml). The same number of these cells produced 1.0 ng of C3/day in the absence
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
of TNF-a and 30 ng of C3/day in the presence of 10 u/ml of TNF-a (Fig. 4), corresponding to 0.0001% and 0.002% of C3 present in 1.0 ml of serum (1300 Ag/ml), respectively. Thus, a relative amount (secreted amount by culture/amount in NHS) of factor D is much higher than that of C3. These findings suggest the possibility that factor D synthesized by gastric epithelium might contribute to the factor D level detectable in circulating blood. Although secreted factor D has a similar molecular weight to factor D found in serum (Fig. 4), factor D activity was not detected in these supernatants when assayed by the method using CVF-B complex [16] (data not shown). Thus, factor D detected in culture supernatants may not be identical to the serum factor D. TNF-a was already shown to enhance C3 production and to suppress C2 production by KATO-III cells [17], when added to the culture medium. This dual effect of TNF-a was also observed with other cell lines tested in the present study (Fig. 3). Since TNF also enhanced factor B production (Fig. 3), gastric epithelial cells may produce C3 and factor B in response to TNF, as reported in the case of fibroblasts [18]. It also appears based on our observations that factor D may be produced constitutively. TNF is reported to participate in inflammatory diseases in gastrointestinal tract. According to Braegger et al. [19], the TNF-a level in stool increases in
847
patients with intestinal inflammation. van Dullemen et al. [20] reported that TNF secretion in the gastrointestinal tract may constitute a major factor in the pathophysiology of Crohn’s disease. On the other hand, in vitro study by Mai et al. [21] showed that Helicobacter pylori activates monocytes, resulting in the secretion of TNF. Besides, C3a and C3a desArg enhance TNF synthesis by adherent monocytes at local inflammatory sites [22]. Thus, TNF may be generated probably by monocytes in gastrointestinal tissue in cases of inflammation. Then, generated TNF may stimulate gastric epithelial cells to synthesize C3 and factor B as shown by the present study. Thus, there may be a possibility of positive feed back or an amplification mechanism (Fig. 5), between TNF generation by monocytes and C3 and factor B secretion by epithelial cells, followed by activation of the complement system generating C3a and C3a desArg in inflammatory gastric tissues. The dual effects of TNF-a on complement secretion by gastric cells may have special significance. Since C2 is produced in the absence of TNF-a, and C3 and factor B are produced in the presence of TNFa, while factor D is synthesized constitutively, there may be another mechanism at work in which proteins of the classical complement pathway are maintained in a physiological state but the alternative complement pathway is activated in the pathological state.
Fig. 5. Proposed mechanism of complement participation in the inflammatory process in gastric membrane. There may be a positive feedback or amplification mechanism between TNF generation and complement synthesis in inflammatory gastric membrane.
848
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
References [1] Morgan BP, Gasque P. Extrahepatic complement biosynthesis: where, when and why? Clin Exp Immunol 1997;107:1 – 7. [2] Colten HR. Tissue-specific regulation of inflammation. J Appl Physiol 1992;72:1 – 7. [3] Kitano E, Kitamura H. Synthesis of the third component complement (C3) by human gastric cancer-derived cell lines. Clin Exp Immunol 1993;94:273 – 8. [4] Naughton MA, Botto M, Carter MJ, Alexander GJM, Goldman JM, Walport MJ. Extrahepatic secreted complement C3 contributes to circulating C3 levels in humans. J Immunol 1996;156:3051 – 6. [5] Vroon DH, Schultz DR, Zarco RM. The separation of nine components and two inactivators of components of complement in human serum. Immunochemistry 1970;7:43 – 61. [6] Tack BF, Morris SC, Prahl JW. Third component of human complement: purification from plasma and physicochemical characterization. Biochemistry 1976;15:4513 – 21. [7] Sekiguchi M, Sakakibara K, Fujii J. Establishment of cultured cell lines derived from a human gastric carcinoma. Jpn J Exp Med 1978;48:61 – 8. [8] Kitano E, Kitamura H. Synthesis of factor D by normal human hepatocytes. Int Arch Allergy Immunol 2000;122:299 – 302. [9] Inai S, Kitamura H, Hiramatsu S, Nagaki K. Deficiency of the ninth component of complement in man. J Clin Lab Immunol 1979;2:85 – 7. [10] Towbin H, Staehelin T, Goldon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci U S A 1979;76:4350 – 4. [11] White RT, Damm D, Hancock N, Rosen BS, Lowell BB, et al. Human adipsin is identical to complement factor D and is expressed at high levels in adipose tissue. Biol Chem 1992; 267:9210 – 3. [12] Choy LN, Rosen BS, Spiegelman BM. Adipsin and an endogeneous pathway of complement from adipose tissue. J Biol Chem 1992;267:12736 – 41. [13] Whaley K. Biosynthesis of the complement components and
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
the regulatory proteins of the alternative complement pathway by human peripheral blood monocytes. J Exp Med 1980; 151:501 – 6. Barnum SR, Volanakis JE. Biosynthesis of complement protein D by HepG2 cells: a comparison of D produced by HepG2 cells, U937 cells and blood monocytes. Eur J Immunol 1985; 15:1148 – 51. Barnum SR, Ishii Y, Agrawal A, Volanakis JE. Production and interferon-g-mediated regulation of complement component C2 and factors B and D by astroglioma cell line U105-MG. Biochem J 1992;287:595 – 601. Miyama A, Kato T, Horai S. Trypsin-activated complex of human factor B with cobra venom factor (CVF), cleaving C3 and C5 and generating a lytic factor for unsensitized guinea pig erythrocytes: I. Generation of the activated complex. Biken J 1975;18:193 – 204. Kitano E, Kitamura H. Dual effects of TNF on synthesis of complement components by a gastric cancer-derived cell line, KATO-III. Int Arch Allergy Immunol 1999;119:54 – 9. Katz Y, Stav D, Barr J, Passwell JH. IL-13 results in differential regulation of the complement proteins C3 and factor B in tumour necrosis factor (TNF)-stimulated fibroblasts. Clin Exp Immunol 1995;101:150 – 6. Braegger CP, Nicholis S, Murch SH, Stephens S, MacDonald TT. Tumor necrosis factor-a in stool as a marker of intestinal inflammation. Lancet 1992;339:89 – 91. van Dullemen HM, van Deventer SJ, Hommes DW, Bijl HA, Jansen J, et al. Treatment of Crohn’s disease with anti-tumor necrosis factor chimeric antibody (cA2). Gastroenterology 1995;109:129 – 35. Mai UEH, Perez GIP, Wahl LM, Wahl SM, Blaser MJ, Smith PD. Soluble surface proteins from Helicobacter pyroli activate monocytes/macrophages by lipopolysaccharide-independent mechanism. J Clin Invest 1991;87:894 – 900. Takabayashi T, Vannier E, Clark BD, Margolis NH, Dinarello CA, et al. A new biologic role for C3a and C3a desArg: regulation of TNF-a and IL-1h synthesis. J Immunol 1996; 156: 3455 – 60.
Synthesis of factor D by gastric cancer-derived cell lines Etsuko Kitano, Hajime Kitamura* Department of Clinical Laboratory Science, Osaka Prefecture College of Health Sciences, Habikino, Osaka, Japan Received 18 June 2001; received in revised form 25 February 2002; accepted 11 March 2002
Abstract Synthesis of complement components in vitro by four human gastric cancer-derived cell lines, MKN28, MKN74, MKN45 and KATO-III, was studied. When these cells were cultured for 3 days without addition of any stimulator, 0.94F0.49, 2.10F0.59, 7.29F5.94 and 2.47F1.34 ng of factor D/106 cells were detected in supernatants of MKN28, MKN74, MKN45 and KATO-III, respectively. Factor D production by these cells was reversibly inhibited by the presence of cycloheximide. Factors B, C3 and C2 were also detected in protein-free culture medium of these cell lines. Addition of tumour necrosis factor (TNF) to culture enhanced C3 and factor B secretion but depressed C2 secretion, without any distinct effect on factor D secretion. Since all cell lines tested secreted significant amounts of factor D without addition of any stimulator in medium, it is possible that factor D may be synthesized by gastric epithelial cells physiologically and constitutively. From a quantitative analysis of factor D secretion by these cells, factor D secreted by gastric tissue is likely to contribute to the factor D level in circulating blood. The possible mechanism of participation of complement system in inflammation of gastric epithelium was proposed. Thus, the present study may be significant for clarification of the mode of extrahepatic complement synthesis participating in mucosal immunity. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Complement; Factor D; Gastric epithelial cell; TNF
1. Introduction A number of studies related to the synthesis of complement components revealed that various kinds of cells other than hepatocytes can also produce complement components [1]. It is now believed that extrahepatic production of complement may contribute to complement levels in plasma and that local
*
Corresponding author. Tel.: +81-729-50-2111x3031; fax: +81-729-50-2126. E-mail address: [email protected] (H. Kitamura).
production of complement is important in homeostasis and immune defense in tissue [2]. We reported that all four kinds of cell lines derived from human gastric cancer tissue synthesize C3 in response to TNF, suggesting physiological or pathological production of C3 by gastric tissue [3]. This finding supports the hypothesis of Naughton et al. [4] that a significant proportion of the total systemic pool of C3 is synthesized extrahepatically and that the epithelium of the gastrointestinal tract is likely to be responsible for C3 of extrahepatic origin. The present study deals with the secretion of factor D as well as factors B, C2 and C3 by human gastric cancer-derived cell lines.
1567-5769/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 7 - 5 7 6 9 ( 0 2 ) 0 0 0 2 8 - 0
844
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
2. Materials and methods 2.1. Reagents and antibodies Reagents and antibodies were obtained as follows: RPMI 1640, MEM and protein-free medium (PFHMII) were from Gibco (Grand Island, NY); fetal calf serum (FCS) was from CC Laboratories (Cleveland, OH); natural human TNF-a was from Hayashibara (Okayama, Japan); IL-1h and IL-6 were from Genzyme (Boston, MA); IFN-g was from Shionogi (Osaka, Japan); LPS was from Paesel (Germany). Antibody to human C3 (goat anti-C3 IgG) was from Cappel (West Chester, PA); rabbit antiserum to human C3c (rabbit anti-C3c) and to human C2 (rabbit anti-C2) were from Nordic, The Netherlands; sheep anti-factor D and sheep anti-factor B were from Binding Site (Birmingham, UK); rabbit anti-factor D was from Wako (Osaka); rabbit anti-factor B was from DAKO (Denmark); horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG and HRPconjugated rabbit anti-goat IgG and HRP-conjugated goat anti-rabbit IgG were from Bio-Rad (Richmond, CA). Sera from 40 healthy donors were combined and used as pooled normal human serum (NHS). Isolated factor D was kindly provided by Dr. K.S. Hong (JCR Pharmaceuticals, Kobe, Japan) and isolated factor B was purchased from Sigma (St. Louis, MO). C2 [5] and C3 [6] were isolated in our laboratory as described earlier. 2.2. Cell lines Four cell lines derived from human gastric cancer, MKN28, MKN45, MKN74 and KATO-III, were provided by The Japanese Cancer Research Resources Bank [3]. The histological origin of KATO-III cells are signet-ring carcinoma [7] and that of the other cell lines are adenocarcinoma [3]. Cells were grown in medium containing RPMI 1640 (45% v/v), MEM (45% v/v) and FCS (10% v/v).
plates (Falcon). These cell cultures were maintained in a humidified atmosphere of 5% CO2/95% air at 37 jC for 1 – 5 days. The culture medium was then harvested, centrifuged to remove cellular debris and maintained at 70 jC prior to study. 2.4. Complement assay ELISA for factors D and C3: The factor D protein [8] and the C3 protein [3] were quantitated by the sandwich ELISA technique. The factor B protein was also quantitated by the sandwich ELISA by using sheep anti-factor B, rabbit anti-factor B and HRPconjugated anti-rabbit IgG. Functional assay for C2: C2 activity was assayed by immune hemolysis as reported previously [9]. 2.5. Immunoblot analysis The culture supernatants, together with isolated component and NHS, were subjected to 2 – 15% SDS-PAGE under non-reducing conditions and transferred onto nitrocellulose according to the method of Towbin et al. [10]. After blocking unreacted sites and washing, blots were developed using the corresponding antibody and HRP-conjugated anti-IgG and the HRP substrate (Konica Immunostaining HRP1000). 2.6. Statistical analysis All data of concentrations of complement components in culture supernatants are expressed as the meanFS.D. of at least two experiments in triplicates. Data were analysed for statistical significance using Friedman test (StatView 5.0). Differences were considered significant for p<0.001.
3. Results 3.1. Factor D secretion by human gastric cancerderived cancer cell lines
2.3. Cell culture conditions To monitor complement secretion, cells were washed and cultured in fresh protein-free medium, with or without cytokine, in sterile 12-well culture
Gastric cancer-derived cell lines were cultured in PFHM-II without addition of any stimulator. Culture supernatants were sampled and assessed for antigenic factor D by ELISA. Fig. 1 clearly shows that these
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
845
trations of TNF-a for 2 days. Culture supernatants were assayed for factor B, factors D and C3 by ELISA and for hemolytic activity of C2. Results are shown in Fig. 3. Statistical analysis by Friedman test clearly showed that TNF-a enhanced factor B secretion and C3 secretion but suppressed C2 secretion by all of these cell lines in a dose-dependent fashion, while TNF-a did not show any distinct effect on factor D secretion. 3.5. Immunoblot analysis of factor D and factor B in culture supernatants
Fig. 1. Factor D secretion by human gastric cancer-derived cell lines. MKN28 cells (D-D), MKN74 cells (E-E), MKN45 cells (.-.) or KATO-III cells (5-5) were cultured in PFHM-II without addition of any cytokine. Culture supernatants were collected after culture for 1 – 5 days and assayed for factor D. Data represents meanFS.D. of two experiments in triplicates.
Culture supernatants of MKN45, MKN74 and KATO-III were applied for immunoblot analysis of factor D and factor B. As shown in Fig. 4, factor D secreted by MKN45, MKN74 or KATO-III exhibited a single band corresponding to a molecular weight similar to factor D in NHS and isolated factor D. Factor B secreted by these gastric cancer-derived cell lines also exhibited a single band of a molecular
cells produced increasing amounts of factor D over a 5-day period ( p<0.001 by Friedman test). 3.2. Effect of cycloheximide on factor D production Factor D secretion by MKN28 and KATO-III cell lines were completely blocked by cycloheximide treatment and was restored after removal of the inhibitor (Fig. 2). Similar results were obtained with other cell lines (not shown). These results indicate that factor D is synthesized de novo. 3.3. Effect of IFN-g, IL-1h and IL-6 on factor D secretion Factor D was assayed in culture supernatants obtained after incubation for 2 days in PFHM-II containing varying amounts of IL-6 (0 –300 ng/ml), IL-1h (0 – 33 ng/ml) or IFN-g (0 – 50,000 u/ml). Neither of these cytokines showed any distinct effect on factor D production (not shown). 3.4. Effect of TNF-a on factor B, factor D, C3 and C2 secretion MKN28 cells, MKN45 cells and KATO-III cells were cultured in PFHM-II containing varying concen-
Fig. 2. Effect of cycloheximide on factor D secretion. MKN28 cells or KATO-III cells were incubated in the presence (n) or absence (n) of cycloheximide (CHX, 1.0 Ag/ml) for 2 days, after which the cells were washed and incubated for another 2 days in the medium without CHX. Culture supernatants were collected and assayed for factor D. Data represents meanFS.D. of two experiments in triplicates.
846
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
Fig. 4. Immunoblot analysis of factor D in culture supernatants. Culture supernatants of KATO-III cells (K), MKN28 cells (28) and MKN74 cells (74), together with isolated factor D and NHS, were electrophoresed on a 2 – 15% SDS-PAGE and blotted onto nitrocellulose paper. The blot was analyzed for factor D.
weight similar to factor B in NHS and isolated factor B (not shown). These observations suggest that factors B and D molecules secreted by these cell lines might be similar to factors B and D in NHS.
4. Discussion
Fig. 3. Effect of TNF-a on the secretion of C3, C2, factor B and factor D. MKN28 cells (D-D), MKN45 cells (.-.) or KATO-III cells (5-5) were cultured in PFHM-II containing varying concentrations of TNF-a for 2 days. Culture supernatants were collected and assayed for factor B, factors D and C3 by ELISA and for hemolytic activity of C2. Data represents meanFS.D. of two experiments in triplicates.
Factor D has been shown to be identical to adipsin, which is synthesized almost exclusively by adipocytes [1,11,12]. However, human monocytes/macrophages [13], HepG2 cells [14] and astrocytes [15] were also reported to synthesize factor D. Recently, we reported that normal human hepatocytes secrete a significant amount of factor D in culture supernatants, suggesting that liver might be a major source of factor D in serum [8]. The present study clearly shows that factor D is secreted by gastric cancer-derived cell lines. This may be the first report of factor D synthesis by gastric epithelial cells. Of interest is that all of the tested cell lines secreted factor D without addition of any stimulator, suggesting the possibility that gastric epithelial cells might produce factor D constitutively. As to the amount of secreted factor D in culture supernatants, we have found that 106 gastric cells produced factor D at the rate of up to 1.0 ng/day (Fig. 1), corresponding to 0.1% of the protein present in 1.0 ml of serum (1.0 –2.0 Ag/ml). The same number of these cells produced 1.0 ng of C3/day in the absence
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
of TNF-a and 30 ng of C3/day in the presence of 10 u/ml of TNF-a (Fig. 4), corresponding to 0.0001% and 0.002% of C3 present in 1.0 ml of serum (1300 Ag/ml), respectively. Thus, a relative amount (secreted amount by culture/amount in NHS) of factor D is much higher than that of C3. These findings suggest the possibility that factor D synthesized by gastric epithelium might contribute to the factor D level detectable in circulating blood. Although secreted factor D has a similar molecular weight to factor D found in serum (Fig. 4), factor D activity was not detected in these supernatants when assayed by the method using CVF-B complex [16] (data not shown). Thus, factor D detected in culture supernatants may not be identical to the serum factor D. TNF-a was already shown to enhance C3 production and to suppress C2 production by KATO-III cells [17], when added to the culture medium. This dual effect of TNF-a was also observed with other cell lines tested in the present study (Fig. 3). Since TNF also enhanced factor B production (Fig. 3), gastric epithelial cells may produce C3 and factor B in response to TNF, as reported in the case of fibroblasts [18]. It also appears based on our observations that factor D may be produced constitutively. TNF is reported to participate in inflammatory diseases in gastrointestinal tract. According to Braegger et al. [19], the TNF-a level in stool increases in
847
patients with intestinal inflammation. van Dullemen et al. [20] reported that TNF secretion in the gastrointestinal tract may constitute a major factor in the pathophysiology of Crohn’s disease. On the other hand, in vitro study by Mai et al. [21] showed that Helicobacter pylori activates monocytes, resulting in the secretion of TNF. Besides, C3a and C3a desArg enhance TNF synthesis by adherent monocytes at local inflammatory sites [22]. Thus, TNF may be generated probably by monocytes in gastrointestinal tissue in cases of inflammation. Then, generated TNF may stimulate gastric epithelial cells to synthesize C3 and factor B as shown by the present study. Thus, there may be a possibility of positive feed back or an amplification mechanism (Fig. 5), between TNF generation by monocytes and C3 and factor B secretion by epithelial cells, followed by activation of the complement system generating C3a and C3a desArg in inflammatory gastric tissues. The dual effects of TNF-a on complement secretion by gastric cells may have special significance. Since C2 is produced in the absence of TNF-a, and C3 and factor B are produced in the presence of TNFa, while factor D is synthesized constitutively, there may be another mechanism at work in which proteins of the classical complement pathway are maintained in a physiological state but the alternative complement pathway is activated in the pathological state.
Fig. 5. Proposed mechanism of complement participation in the inflammatory process in gastric membrane. There may be a positive feedback or amplification mechanism between TNF generation and complement synthesis in inflammatory gastric membrane.
848
E. Kitano, H. Kitamura / International Immunopharmacology 2 (2002) 843–848
References [1] Morgan BP, Gasque P. Extrahepatic complement biosynthesis: where, when and why? Clin Exp Immunol 1997;107:1 – 7. [2] Colten HR. Tissue-specific regulation of inflammation. J Appl Physiol 1992;72:1 – 7. [3] Kitano E, Kitamura H. Synthesis of the third component complement (C3) by human gastric cancer-derived cell lines. Clin Exp Immunol 1993;94:273 – 8. [4] Naughton MA, Botto M, Carter MJ, Alexander GJM, Goldman JM, Walport MJ. Extrahepatic secreted complement C3 contributes to circulating C3 levels in humans. J Immunol 1996;156:3051 – 6. [5] Vroon DH, Schultz DR, Zarco RM. The separation of nine components and two inactivators of components of complement in human serum. Immunochemistry 1970;7:43 – 61. [6] Tack BF, Morris SC, Prahl JW. Third component of human complement: purification from plasma and physicochemical characterization. Biochemistry 1976;15:4513 – 21. [7] Sekiguchi M, Sakakibara K, Fujii J. Establishment of cultured cell lines derived from a human gastric carcinoma. Jpn J Exp Med 1978;48:61 – 8. [8] Kitano E, Kitamura H. Synthesis of factor D by normal human hepatocytes. Int Arch Allergy Immunol 2000;122:299 – 302. [9] Inai S, Kitamura H, Hiramatsu S, Nagaki K. Deficiency of the ninth component of complement in man. J Clin Lab Immunol 1979;2:85 – 7. [10] Towbin H, Staehelin T, Goldon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci U S A 1979;76:4350 – 4. [11] White RT, Damm D, Hancock N, Rosen BS, Lowell BB, et al. Human adipsin is identical to complement factor D and is expressed at high levels in adipose tissue. Biol Chem 1992; 267:9210 – 3. [12] Choy LN, Rosen BS, Spiegelman BM. Adipsin and an endogeneous pathway of complement from adipose tissue. J Biol Chem 1992;267:12736 – 41. [13] Whaley K. Biosynthesis of the complement components and
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
the regulatory proteins of the alternative complement pathway by human peripheral blood monocytes. J Exp Med 1980; 151:501 – 6. Barnum SR, Volanakis JE. Biosynthesis of complement protein D by HepG2 cells: a comparison of D produced by HepG2 cells, U937 cells and blood monocytes. Eur J Immunol 1985; 15:1148 – 51. Barnum SR, Ishii Y, Agrawal A, Volanakis JE. Production and interferon-g-mediated regulation of complement component C2 and factors B and D by astroglioma cell line U105-MG. Biochem J 1992;287:595 – 601. Miyama A, Kato T, Horai S. Trypsin-activated complex of human factor B with cobra venom factor (CVF), cleaving C3 and C5 and generating a lytic factor for unsensitized guinea pig erythrocytes: I. Generation of the activated complex. Biken J 1975;18:193 – 204. Kitano E, Kitamura H. Dual effects of TNF on synthesis of complement components by a gastric cancer-derived cell line, KATO-III. Int Arch Allergy Immunol 1999;119:54 – 9. Katz Y, Stav D, Barr J, Passwell JH. IL-13 results in differential regulation of the complement proteins C3 and factor B in tumour necrosis factor (TNF)-stimulated fibroblasts. Clin Exp Immunol 1995;101:150 – 6. Braegger CP, Nicholis S, Murch SH, Stephens S, MacDonald TT. Tumor necrosis factor-a in stool as a marker of intestinal inflammation. Lancet 1992;339:89 – 91. van Dullemen HM, van Deventer SJ, Hommes DW, Bijl HA, Jansen J, et al. Treatment of Crohn’s disease with anti-tumor necrosis factor chimeric antibody (cA2). Gastroenterology 1995;109:129 – 35. Mai UEH, Perez GIP, Wahl LM, Wahl SM, Blaser MJ, Smith PD. Soluble surface proteins from Helicobacter pyroli activate monocytes/macrophages by lipopolysaccharide-independent mechanism. J Clin Invest 1991;87:894 – 900. Takabayashi T, Vannier E, Clark BD, Margolis NH, Dinarello CA, et al. A new biologic role for C3a and C3a desArg: regulation of TNF-a and IL-1h synthesis. J Immunol 1996; 156: 3455 – 60.