British /ournu1 OJ Hacmutologg. 199 I , 78. 469-47 3
Autocrine and/or paracrine mechanism operate during the growth of human bone marrow fibroblasts A K I R OK I M U R A , OSAMU K A T O H , HIDEOHYODO,S H I Z U Y O KUSUMI A N D ATSUSHIKURAMOTO Department of Internal Medicine, Research lnstitute for Nuclear Medicine and Biology, Hiroshima University
Received 4 March 1991; accepted for publication 27 March 1991
Summary. The growth of human marrow fibroblasts was studied at various plating cell densities. The growth rate of fibroblasts cultured with human plasma derived serum (PDS) increased depending on the cell density plated, whereas fibroblasts cultured with human serum proliferated almost independent of the cell density. Conditioned medium obtained from the culture with PDS at high cell density enhanced the growth of fibroblasts plated at low density, as
well as that of fibroblast colony forming cells. Characterization of this conditioned medium showed heat unstable and acid stable peptide factor(s) with high molecular weight ( - loh)may be responsible for this activity. These results suggest that autocrine and/or paracrine mechanism can operate during the proliferation of marrow fibroblasts, a process thought to be involved in the progression of marrow fibrosis.
Myelofibrosis is common in patients with myeloproliferative disorders (MPD): however, its pathogenesis has not yet been clarified (McCarthy. 1985). It has been suggested that abnormal leakage or release from megakaryocytes of some growth factors, like platelet derived growth factor (PDGF). epidermal growth factor (EGF) and transforming growth factor-beta (TGF-beta). induces the proliferation of marrow fibroblasts (Castro-Malaspina et al. 1981: Katoh et d . 1988: Kimura et al, 1988a. 1989: Rosenfield et al. 1985). In addition, we have recently shown that marrow fibroblasts from MPD patients show increased sensitivity to human serum mitogens (Kimura rt al, 1988b). Although these mechanisms are important for the initiation as well as progression of marrow fibrosis, many other mechanisms are thought to be operating. In this paper we tested the possibility that autocrine and/or paracrine mechanisms operate during the proliferation of marrow fibroblasts in vitro.
6 0 0 g. Low density mononuclear cells ( 5 x loh)were separated. suspended in alpha minimum essential medium (alphaMEM) supplemented with 15% fetal calf serum (FCS). and incubated at 3 7°C in 5% C02-9 5% air. After 4 h incubation the non-adherent cells were washed out and incubation was continued for about 3 weeks with medium being changed once every week. When the culture reached confluency. cells were subcultured with the same medium. The resulting cultured fibroblasts between the third and fifth passage were used in the following experiments. The fibroblasts stained positively with alkaline phosphatase, acid phosphatase and PAS. but were negative on Sudan black staining. They were nonphagocytic when using iron particles. They did not react with antibodies specific to macrophage (CD68. DAKO). By these criteria, purity of the passaged fibroblasts were more than 99%. Proliferation offibroblasts To compare the growth rate of fibroblasts in the presence of 5% plasma derived serum (PDS) (Ross et al. 1978) and 5% human serum (HS) which was formed by the clotting of whole blood, passaged cells were plated in 2 cm2 plastic wells at the density of 100-5000 cells per cm2and cultured at 3 7°C. Cell number was counted every day, either fixing cells with methanol and staining by the May-Giernsa method when cell number was small, or after detaching cells with trypsin-EDTA when cell number was greater than 1000. Each experiment was done in triplicate. For each culture, doubling time was calculated from the growth curve, and growth rate was expressed as per cent doubling time of that obtained at the lowest cell density ( 100 cells per cm2) in 5% PDS culture.
MATERIALS A N D METHODS Cell culture. Bone marrow aspirates anticoagulated with heparin were obtained from subjects with normal marrow. Ages of the donors were between 38 and 52 years. Hone marrow cells were centrifuged at 90 g for 1 5 min to exclude platelet-rich plasma, and the marrow cell rich fraction was further centrifuged on a Ficoll-Hypaque density gradient a t Correspondence: Dr Akiro Kirnura. Department of Internal Medicine, Research Institute for Nuclear Medicine and Biology, Hiroshima Ilniversity. Kasurni 102-3. Minanii-ku. Hiroshima 734. Japan.
469
470
Akiro Kimura et al
To see the effect of a conditioned medium on the fibroblast proliferation. passaged cells were plated in 2 cm2plastic wells at 100 cells per cmz, and incubated for 5 d with alpha-MEM5% PDS in the presence of 20% of the conditioned medium obtained as described below. Cell number was then counted microscopically after fixation and staining by the MayGiemsa method. To test the inhibition of the conditioned medium dependent growth by antibodies to growth factors, cells were plated and incubated under the same conditions with antibody against either PDGF A chain, PDGF B chain, interleukin l a (ILla),IL-lP or tumour necrosis factor (TNF). To test fibroblast growth promoting activity of fractionated conditioned medium, passaged cells were plated in 2 cm2 plastic wells at 500 cells per cmL,and incubated with alphaMEM-2% PDS for 1 d. 15% of fractionated conditioned medium was then added, and the incubation was further continued for another 2 4 h. 12 h before the end of incubation ’H-thymidine (1 8.5 kBq/ml) was added, and its incorporation by the cells was determined according to Kimura et al (1 988a). The growth promoting activity was expressed as the ratio of ’H-incorporation with test sample to that without sample (stimulation index). The mean ’H-incorporation without sample was 890,cpm. Fibroblast colony formation. Marrow low density mononuclear cells (2 x lo5)obtained as described above were washed, suspended in 2 ml of alpha MEM-SX PDS, added to 35 mm plastic dishes and incubated for 4 h. Non-adherent cells were removed by washing, and incubation was continued in the presence or absence of conditioned medium. After 7 d the cultures were fixed, stained and the number of fibroblast colonies was counted. A fibroblast colony was defined as a cell mass which consisted of more than 1 1 fibroblasts. Conditioned medium. Fibroblasts between the third and fifth passage were plated at 5000 cells per cm2 and cultured with alpha-MEM-5% PDS for 5 d. The supernatant was collected and stored at -20°C until use. To characterize some of the properties of factor(s) in the conditioned medium, 0.5 ml of conditioned medium was subjected to heating at various temperatures for between 1 and 2 4 h. An additional 0.5 ml aliquot was exposed to an equal volume of 1 M acetic acid and allowed to stand for 30 min at room temperature. The medium was then neutralized and lyophilized prior to reconstitution in 0 . 5 ml water. An additional aliquot of medium was exposed to 10 pg/ml of trypsin, incubated for 3 h at 37OC and neutralized with 2 0 pg/ml soybean trypsin inhibitor. These treated samples were dialysed against phosphate-buffered saline (PBS) (pH 7.4). To fractionate the conditioned medium into four portions. the conditioned medium was dialysed against PBS and subjected to ultrafiltration using ultrafiltration membranes, Y M I O , Y M l 0 O and XM300 (Amicon, U.S.A.). sequentially. YMI 0, YMI 00 and XM300 ultrafiltrate macromolecules with molecular weight less than l o x 10’. lOOx 10’ and 3 0 0 x 1 0 ’ . respectively. In each step the filtered solution was collected, and the concentrated conditioned medium was diluted back to the initial volume with PBS. To further analyse the XM 300 concentrate, the concentrate was applied to a Sephacryl S-300 column (Pharmacia) and eluted
with PBS. Each fraction (0.35 ml) was collected and assayed for fibroblast growth promoting activity. Growth factors and antibodies. Purified human PDGF and rabbit anti PDGF A- and B-chain specific antibodies were kindly supplied by Dr Deuel (Washington University, St Louis) and by Dr Hart (Zymogenetics. Seattle), respectively. FCS was purchased from MAB company (U.S.A.). Recombinant human interleukin 1 (rhIL-I) and anti IL-1 antibody were supplied by Dainihon Pharmaceutical Co.. rhIL-1 and anti IL-1 by Otsuka Pharmaceutical Co., and rh TNF and anti TNF by Asahi Chemical Industry. RESULTS The growth rates of marrow fibroblasts were compared in the presence of either 5% HS or 5% PDS. Fig 1 shows a typical growth pattern of fibroblasts from one donor at various plating cell densities. In the presence of PDS the growth rate increased depending on the cell density plated. In contrast. in the presence of HS the rate was higher than that with PDS in each plating cell density although the difference was not significant at two higher plating cell densities, and did not change even when cell density was increased. From the data of growth curves for different donors, the doubling time was calculated according to the cell density. The doubling time differed among donors mainly depending on the donor age. However, when the doubling time at the plating cell density of 100 cells per cmz in 5% PDS culture from each donor was defined as loo%, the per cent doubling time was not so variable among donors. Fig 2 shows that the per cent doubling time ofthe fibroblast growth with 5% PDS decreased as plating cell density increased. On the other hand, with 5% HS. the mean per cent doubling time at the lowest cell density
I0000
- 5000
N
E
Y
v
& 1000
n
5
500
C
Q,
V
I00
0
I
2
3
4 (day)
Fig 1. Growth curve of marrow fibroblasts in relation to plating cell density. Various numbers of marrow fibroblasts from one donor were or 5% HS ( O ) ,and cell plated at day 0,incubated either 5% PDS (0) numbers were counted each day.
Autocririe Growth of Marrow Fibroblasts
':3
471
Table 1. Stability of growth promoting activity of marrow fibroblast conditioned medium. Fibroblast conditioned medium from one donor was treated as presented. Each treated conditioned medium was tested for their growth-promoting activity as described in Materials and Methods. The residual activity was expressed as per cent ofcontrol (untreated conditioned medium).
8
2
% ' ,,
,
100
1000
10000
Control Dialysis 37'C. 24 h 56°C. 24 h 90OC. 1 h Acid treatment Trypsin treatment
Cell number plated (/crn2) Fig 2 . Effect of plating cell density on the doubling time of marrow fibroblast growth. Fibroblasts at various plating cell densities were cultured in the presence of either 5(Xj PIX ( 0 )or 5% HS ( 0 ) .The doubling times were calculated from the growth curves. and expressed as per cent doubling time. The doubling times of PDS culture at cell density of IOO/cm' (control) were 7 . 7 f 1.9 d (m+SI>). The data from four donors are shown. Vertical line indicates mean+one standard deviation. * P < 0 . 0 5 and **P
Autocrine and/or paracrine mechanism operate during the growth of human bone marrow fibroblasts A K I R OK I M U R A , OSAMU K A T O H , HIDEOHYODO,S H I Z U Y O KUSUMI A N D ATSUSHIKURAMOTO Department of Internal Medicine, Research lnstitute for Nuclear Medicine and Biology, Hiroshima University
Received 4 March 1991; accepted for publication 27 March 1991
Summary. The growth of human marrow fibroblasts was studied at various plating cell densities. The growth rate of fibroblasts cultured with human plasma derived serum (PDS) increased depending on the cell density plated, whereas fibroblasts cultured with human serum proliferated almost independent of the cell density. Conditioned medium obtained from the culture with PDS at high cell density enhanced the growth of fibroblasts plated at low density, as
well as that of fibroblast colony forming cells. Characterization of this conditioned medium showed heat unstable and acid stable peptide factor(s) with high molecular weight ( - loh)may be responsible for this activity. These results suggest that autocrine and/or paracrine mechanism can operate during the proliferation of marrow fibroblasts, a process thought to be involved in the progression of marrow fibrosis.
Myelofibrosis is common in patients with myeloproliferative disorders (MPD): however, its pathogenesis has not yet been clarified (McCarthy. 1985). It has been suggested that abnormal leakage or release from megakaryocytes of some growth factors, like platelet derived growth factor (PDGF). epidermal growth factor (EGF) and transforming growth factor-beta (TGF-beta). induces the proliferation of marrow fibroblasts (Castro-Malaspina et al. 1981: Katoh et d . 1988: Kimura et al, 1988a. 1989: Rosenfield et al. 1985). In addition, we have recently shown that marrow fibroblasts from MPD patients show increased sensitivity to human serum mitogens (Kimura rt al, 1988b). Although these mechanisms are important for the initiation as well as progression of marrow fibrosis, many other mechanisms are thought to be operating. In this paper we tested the possibility that autocrine and/or paracrine mechanisms operate during the proliferation of marrow fibroblasts in vitro.
6 0 0 g. Low density mononuclear cells ( 5 x loh)were separated. suspended in alpha minimum essential medium (alphaMEM) supplemented with 15% fetal calf serum (FCS). and incubated at 3 7°C in 5% C02-9 5% air. After 4 h incubation the non-adherent cells were washed out and incubation was continued for about 3 weeks with medium being changed once every week. When the culture reached confluency. cells were subcultured with the same medium. The resulting cultured fibroblasts between the third and fifth passage were used in the following experiments. The fibroblasts stained positively with alkaline phosphatase, acid phosphatase and PAS. but were negative on Sudan black staining. They were nonphagocytic when using iron particles. They did not react with antibodies specific to macrophage (CD68. DAKO). By these criteria, purity of the passaged fibroblasts were more than 99%. Proliferation offibroblasts To compare the growth rate of fibroblasts in the presence of 5% plasma derived serum (PDS) (Ross et al. 1978) and 5% human serum (HS) which was formed by the clotting of whole blood, passaged cells were plated in 2 cm2 plastic wells at the density of 100-5000 cells per cm2and cultured at 3 7°C. Cell number was counted every day, either fixing cells with methanol and staining by the May-Giernsa method when cell number was small, or after detaching cells with trypsin-EDTA when cell number was greater than 1000. Each experiment was done in triplicate. For each culture, doubling time was calculated from the growth curve, and growth rate was expressed as per cent doubling time of that obtained at the lowest cell density ( 100 cells per cm2) in 5% PDS culture.
MATERIALS A N D METHODS Cell culture. Bone marrow aspirates anticoagulated with heparin were obtained from subjects with normal marrow. Ages of the donors were between 38 and 52 years. Hone marrow cells were centrifuged at 90 g for 1 5 min to exclude platelet-rich plasma, and the marrow cell rich fraction was further centrifuged on a Ficoll-Hypaque density gradient a t Correspondence: Dr Akiro Kirnura. Department of Internal Medicine, Research Institute for Nuclear Medicine and Biology, Hiroshima Ilniversity. Kasurni 102-3. Minanii-ku. Hiroshima 734. Japan.
469
470
Akiro Kimura et al
To see the effect of a conditioned medium on the fibroblast proliferation. passaged cells were plated in 2 cm2plastic wells at 100 cells per cmz, and incubated for 5 d with alpha-MEM5% PDS in the presence of 20% of the conditioned medium obtained as described below. Cell number was then counted microscopically after fixation and staining by the MayGiemsa method. To test the inhibition of the conditioned medium dependent growth by antibodies to growth factors, cells were plated and incubated under the same conditions with antibody against either PDGF A chain, PDGF B chain, interleukin l a (ILla),IL-lP or tumour necrosis factor (TNF). To test fibroblast growth promoting activity of fractionated conditioned medium, passaged cells were plated in 2 cm2 plastic wells at 500 cells per cmL,and incubated with alphaMEM-2% PDS for 1 d. 15% of fractionated conditioned medium was then added, and the incubation was further continued for another 2 4 h. 12 h before the end of incubation ’H-thymidine (1 8.5 kBq/ml) was added, and its incorporation by the cells was determined according to Kimura et al (1 988a). The growth promoting activity was expressed as the ratio of ’H-incorporation with test sample to that without sample (stimulation index). The mean ’H-incorporation without sample was 890,cpm. Fibroblast colony formation. Marrow low density mononuclear cells (2 x lo5)obtained as described above were washed, suspended in 2 ml of alpha MEM-SX PDS, added to 35 mm plastic dishes and incubated for 4 h. Non-adherent cells were removed by washing, and incubation was continued in the presence or absence of conditioned medium. After 7 d the cultures were fixed, stained and the number of fibroblast colonies was counted. A fibroblast colony was defined as a cell mass which consisted of more than 1 1 fibroblasts. Conditioned medium. Fibroblasts between the third and fifth passage were plated at 5000 cells per cm2 and cultured with alpha-MEM-5% PDS for 5 d. The supernatant was collected and stored at -20°C until use. To characterize some of the properties of factor(s) in the conditioned medium, 0.5 ml of conditioned medium was subjected to heating at various temperatures for between 1 and 2 4 h. An additional 0.5 ml aliquot was exposed to an equal volume of 1 M acetic acid and allowed to stand for 30 min at room temperature. The medium was then neutralized and lyophilized prior to reconstitution in 0 . 5 ml water. An additional aliquot of medium was exposed to 10 pg/ml of trypsin, incubated for 3 h at 37OC and neutralized with 2 0 pg/ml soybean trypsin inhibitor. These treated samples were dialysed against phosphate-buffered saline (PBS) (pH 7.4). To fractionate the conditioned medium into four portions. the conditioned medium was dialysed against PBS and subjected to ultrafiltration using ultrafiltration membranes, Y M I O , Y M l 0 O and XM300 (Amicon, U.S.A.). sequentially. YMI 0, YMI 00 and XM300 ultrafiltrate macromolecules with molecular weight less than l o x 10’. lOOx 10’ and 3 0 0 x 1 0 ’ . respectively. In each step the filtered solution was collected, and the concentrated conditioned medium was diluted back to the initial volume with PBS. To further analyse the XM 300 concentrate, the concentrate was applied to a Sephacryl S-300 column (Pharmacia) and eluted
with PBS. Each fraction (0.35 ml) was collected and assayed for fibroblast growth promoting activity. Growth factors and antibodies. Purified human PDGF and rabbit anti PDGF A- and B-chain specific antibodies were kindly supplied by Dr Deuel (Washington University, St Louis) and by Dr Hart (Zymogenetics. Seattle), respectively. FCS was purchased from MAB company (U.S.A.). Recombinant human interleukin 1 (rhIL-I) and anti IL-1 antibody were supplied by Dainihon Pharmaceutical Co.. rhIL-1 and anti IL-1 by Otsuka Pharmaceutical Co., and rh TNF and anti TNF by Asahi Chemical Industry. RESULTS The growth rates of marrow fibroblasts were compared in the presence of either 5% HS or 5% PDS. Fig 1 shows a typical growth pattern of fibroblasts from one donor at various plating cell densities. In the presence of PDS the growth rate increased depending on the cell density plated. In contrast. in the presence of HS the rate was higher than that with PDS in each plating cell density although the difference was not significant at two higher plating cell densities, and did not change even when cell density was increased. From the data of growth curves for different donors, the doubling time was calculated according to the cell density. The doubling time differed among donors mainly depending on the donor age. However, when the doubling time at the plating cell density of 100 cells per cmz in 5% PDS culture from each donor was defined as loo%, the per cent doubling time was not so variable among donors. Fig 2 shows that the per cent doubling time ofthe fibroblast growth with 5% PDS decreased as plating cell density increased. On the other hand, with 5% HS. the mean per cent doubling time at the lowest cell density
I0000
- 5000
N
E
Y
v
& 1000
n
5
500
C
Q,
V
I00
0
I
2
3
4 (day)
Fig 1. Growth curve of marrow fibroblasts in relation to plating cell density. Various numbers of marrow fibroblasts from one donor were or 5% HS ( O ) ,and cell plated at day 0,incubated either 5% PDS (0) numbers were counted each day.
Autocririe Growth of Marrow Fibroblasts
':3
471
Table 1. Stability of growth promoting activity of marrow fibroblast conditioned medium. Fibroblast conditioned medium from one donor was treated as presented. Each treated conditioned medium was tested for their growth-promoting activity as described in Materials and Methods. The residual activity was expressed as per cent ofcontrol (untreated conditioned medium).
8
2
% ' ,,
,
100
1000
10000
Control Dialysis 37'C. 24 h 56°C. 24 h 90OC. 1 h Acid treatment Trypsin treatment
Cell number plated (/crn2) Fig 2 . Effect of plating cell density on the doubling time of marrow fibroblast growth. Fibroblasts at various plating cell densities were cultured in the presence of either 5(Xj PIX ( 0 )or 5% HS ( 0 ) .The doubling times were calculated from the growth curves. and expressed as per cent doubling time. The doubling times of PDS culture at cell density of IOO/cm' (control) were 7 . 7 f 1.9 d (m+SI>). The data from four donors are shown. Vertical line indicates mean+one standard deviation. * P < 0 . 0 5 and **P
