Regulation of DNA synthesis in chicken adipocyte precursor cells by insulin-like growth factors, platelet-derived growth factor and transforming growth factor-\g=b\ S. C. Butterwith and C. Goddard Department of Cellular and Molecular Biology, AFRC Institute of Animal Physiology and Genetics Research, Edinburgh Research Station, Roslin, Midlothian EH25 9ps received
15 March 1991
ABSTRACT
Adipose tissue growth can occur by both hypertrophy and hyperplasia. The capacity for adipocyte hyperplasia in vivo resides in a population of fibroblast-like adipocyte precursor cells but the regulation of the proliferation of these cells by growth factors has not been well characterized. This study was designed to determine the effects of the insulin-like growth factors (IGF-I and IGF-II), platelet-derived growth factor (PDGF) and transforming growth factor-\g=b\1(TGF\x=req-\ \g=b\1)added alone or together on the proliferation of primary adipocyte precursor cells in vitro. Adipocyte precursor cell proliferation measured by [3H]thymidine incorporation into DNA was stimulated by all of these growth factors and was particularly marked with PDGF. IGF-I or IGF-II added together with TGF-\g=b\1produced a greater than additive response and the effect of PDGF was synergistic with that of
of PDGF but the evidence we have suggests that this is unlikely in chicken adipocyte precursors. DNA synthesis in response to TGF-\g=b\1required only a short exposure to the peptide, and conditioned medium from chicken adipocyte precursor cells previously exposed to TGF-\g=b\had no effect on DNA synthesis when added to fresh batches of cells. Addition of TGF-\g=b\1together with PDGF produced a synergistic effect whereas an additive effect would be expected if PDGF mediated the effect of TGF-\g=b\1. IGF-I mRNA is expressed in the Ob 1771 preadipocyte cell line during differentiation, in stromalvascular cells from adipose tissue, and TGF-\g=b\mRNA is expressed in both proliferating and differentiating 3T3-L1 preadipocytes. Together with the data presented here, this would indicate that these peptides have a role in adipocyte development by an autocrine or paracrine mechanism although the source of PDGF in vivo is at present unknown. Journal of Endocrinology (1991) 131, 203\p=n-\209
INTRODUCTION
& Ailhaud, 1987) and proliferation of 3T3-L1 cells (Smith, Wise, Berkowitz et al. 1988). IGF-I mRNA is also expressed in differentiated Ob 1771 preadipocytes, in murine adipose tissue stromal-vascular cells (Doglio, Dani, Fredrikson et al. 1987) and in differentiated primary cultures of porcine adipocyte precursors (Gaskins, Kim, Wright et al. 1990). Transforming growth factor-ß (TGF-ß) also regu¬ lates differentiation of a number of cell types, in¬ cluding adipocytes (Ignotz & Massague, 1985; Torti, Torti, Larrick & Ringold, 1989), but the proliferative response to TGF-ß is dependent on the cell type and can be inhibitory or stimulatory (Moses, Tucker, Leof et al. 1985). Stimulation of DNA synthesis may be an indirect effect of TGF-ß via the production of platelet-
IGF-I at certain concentrations. Stimulation of proliferation of some cell types by TGF-\g=b\has been linked to the secondary production
development of preadipocyte cell lines and cul¬ primary adipocyte precursors has aided the greatly study of adipocyte development and The
ture methods for
has enabled details of the factors which specifically regulate proliferation and differentiation to emerge. Most studies have concentrated on the regulation of differentiation in specific cell lines such as 3T3-L1 and Ob 1771 and less is known about factors which regulate proliferation in primary cultures of adipocyte precursor cells. It has been shown that insulin-like growth factor-I (IGF-I) can stimulate proliferation and differen¬ tiation of rat adipocyte precursor cells (Deslex, Negrel
derived growth factor (PDGF)-like proteins coded by the c-sis proto-oncogene (Leof, Proper, Goustin et al. 1986; Soma & Grotendorst, 1989) which then stimu¬ late proliferation. TGF-ß mRNA is also expressed in proliferating and differentiating 3T3-L1 preadipocytes (Weiner, Shah, Smith et al. 1989). This evidence is consistent with the idea that these growth factors may have an autocrine/paracrine role in adipocyte development but most studies have investigated the role of growth factors added singly to cells in culture, whereas cells in vivo are exposed to a combination of factors. This study was designed to investigate the relationship between IGF peptides, PDGF and TGF-ßl on the regulation of DNA syn¬ thesis in primary cultures of adipocyte precursor cells. MATERIALS AND METHODS
TGF-ßl was obtained from British Biotechnology, Cowley, Oxon, U.K. or was a gift from Dr R. Derynk (Genentech, South San Francisco, CA, U.S.A.). Both batches were equipotent. Recombinant cDNA-
derived human IGF-I, IGF-II and PDGF-AB were from Bachern, Saffron Waiden, Essex, U.K. Bovine serum albumin (BSA, radioimmunoassay grade) was from Sigma, Poole, Dorset, U.K. Dulbecco's A phosphate-buffered saline (PBS) was prepared from tablets according to the manufacturer's instructions
(Oxoid Co., Basingstoke, Hants, U.K.). Trypsin (2-5%, w/v), gentamicin and Ultroser were from Gibco-BRL, Paisley, Strathclyde, U.K. Tissue culture plates (Costar; 24-multiwell plates) and medium 199 (Ml99) were from Northumbria Biologicals, Cramlington, Northumberland, U.K. 6-[3H] Thymidine (25 Ci/mmol) was from Amersham Inter¬ national pic, Amersham, Bucks, U.K. Animals
Female broiler chickens of a commercial line were obtained from D. B. Marshall, Newbridge, Midlothian, U.K. They were housed in brooders with a lighting pattern of 14 h light : 10 h darkness and fed on stan¬ dard broiler diets. Birds were killed at 2-3 weeks of age by cervical dislocation, and abdominal adipose tissue was removed for the preparation of adipocyte
precursor cells.
Cell culture
Adipocyte precursor cells were prepared and cultured in supplemented M199 according to the method of Butterwith & Griffin (1989). They were plated in multiwell plates at a density of 32 000 cells/well
(2 cm2) in 1ml
M199 containing 4% (v/v) Ultroser and incubated for 66 h at 37 °C in a humidified atmos¬ phere containing 5% C02. They were washed and incubated in M199 for a further 24 h before addition of growth factor together with [3H]thymidine (0-25 pCi/well). After incubation for 18 h the super¬ natant was aspirated and the cell monolayer washed with PBS (1 ml) and treated with trypsin (400 pi; 1%, w/v) and NaOH (100 pi; 5 mol/1). The cell suspension was transferred to a microfuge tube, 100 pi BSA (1 %, w/v) and 350 pi perchloric acid (2 mol/1) were added and the samples incubated overnight at 4 °C. After centrifugation at 2800 # for 10 min, the precipitate was washed with ice-cold perchloric acid (1 mol/1) before being resuspended in 300 pi perchloric acid (1 mol/1) and hydrolysis at 70 °C for 30 min. After further centrifugation, a sample of the supernatant was counted for radioactivity in Optiphase X (Beckman
Instruments, High Wycombe, Bucks, U.K.) using
a
Wallac 1410 scintillation counter (Pharmacia LKB Biotechnology, Milton Keynes, Bucks, U.K.). Initial experiments demonstrated that there was a good correlation between [3H]thymidine incorporation and cell number. In order to investigate the time required to observe effects of TGF-ßl, cells were preincubated with TGFßl for 3 h. The medium was removed and cells were washed with fresh medium (1 ml) and then incubated for 18 h in M199 containing [3H]thymidine and IGF-I. Control incubations received no TGF-ßl during preincubation, which was followed by the addition of IGF-I, TGF-ßl, both or medium alone. The pre¬ incubation period was also extended to 4 and 5 h. In a further experiment, the effect of medium con¬ ditioned by cells previously exposed to TGF-ßl was examined. Cells were preincubated with TGF-ßl (1 pg/1) for 6h, washed with fresh medium (1ml) and then incubated for 18 h in M199 alone. This conditioned medium was added to a different batch of growth-restricted cells and [3H]thymidine incor¬ poration into DNA was measured as previously described. Medium conditioned by cells which had not been exposed to TGF-ßl was used as a control and the effect of IGF-I (5 pg/1) alone and a combi¬ nation of IGF-I and TGF-ßl were used as positive controls in each case. Statistics
The ALLFIT program (De Lean, Munson & Rodbard, 1978) was used to analyse dose-response curves and calculate half-maximum effective concentrations for thymidine incorporation (ED50). A two-way analy¬ sis of variance was used to test for synergistic inter¬ actions using the MINITAB statistical package on an Apple Macintosh SE/30 microcomputer. Other data
were
analysed by Student's
statistical
/-test
package.
the
using
1200
same
()
1000
RESULTS
800
The cells were identified as adipocyte precursors on the basis of their ability to differentiate in culture on reaching confluence. Differentiation was character¬ ized by an increase in lipoprotein lipase activity and lipogenesis (Butterwith & Griffin, 1989) and by the accumulation of lipid droplets.
600
W
*'
400 J
o.
1200
•
1 IGF-1
(b)
synthesis incorporation of [3H]thymidine into DNA by adipocyte precursor cells in the absence of added
800
(u.g/1)
.»
The
600
Ultroser or exogenous growth factor varied between experiments but was in the range 500-1800 c.p.m./well with a mean value of 892 + 511 (s.d.; 6) for the addition of M199 here. The experiments reported
8
400
=
containing 4% Ultroser increased the [3H]thymidine incorporation to 75 129+19 789 c.p.m./well (n 5),
4000
1 IGF-II
0
(c)
=
an
84-fold stimulation.
Effect of IGF-I, IGF-II and
3000
The effects of IGF-I, IGF-II and TGF-ßl on [3H] thymidine incorporation into DNA by adipocyte pre¬ cursors when added alone are shown in Fig. 1. Both IGF-I and IGF-II stimulated [3H]thymidine incorpor¬ ation in a dose-dependent manner, with a maximum response (about twofold) at 5 pg/1 with an ED50 of 2-06+1-47 and 0-92 + 0-75 (s.e.m.; d.f. 8 and 7) pg/1 for IGF-I and IGF-II respectively. The ED50 values were not significantly different as determined by a =
P
=
0-54
on
16d.f.)
TGF-ßl also stimulated [3H]thymidine incorpor¬ ation by about threefold with the maximum at con¬ centrations of 1 pg/1 and an ED50 of 0-34+ 0025 (s.e.m.; d.f. 10) pg/1. The response to TGF-ßl was more sensitive than the IGF peptides when the ED50 values were compared (F=4-89, P
15 March 1991
ABSTRACT
Adipose tissue growth can occur by both hypertrophy and hyperplasia. The capacity for adipocyte hyperplasia in vivo resides in a population of fibroblast-like adipocyte precursor cells but the regulation of the proliferation of these cells by growth factors has not been well characterized. This study was designed to determine the effects of the insulin-like growth factors (IGF-I and IGF-II), platelet-derived growth factor (PDGF) and transforming growth factor-\g=b\1(TGF\x=req-\ \g=b\1)added alone or together on the proliferation of primary adipocyte precursor cells in vitro. Adipocyte precursor cell proliferation measured by [3H]thymidine incorporation into DNA was stimulated by all of these growth factors and was particularly marked with PDGF. IGF-I or IGF-II added together with TGF-\g=b\1produced a greater than additive response and the effect of PDGF was synergistic with that of
of PDGF but the evidence we have suggests that this is unlikely in chicken adipocyte precursors. DNA synthesis in response to TGF-\g=b\1required only a short exposure to the peptide, and conditioned medium from chicken adipocyte precursor cells previously exposed to TGF-\g=b\had no effect on DNA synthesis when added to fresh batches of cells. Addition of TGF-\g=b\1together with PDGF produced a synergistic effect whereas an additive effect would be expected if PDGF mediated the effect of TGF-\g=b\1. IGF-I mRNA is expressed in the Ob 1771 preadipocyte cell line during differentiation, in stromalvascular cells from adipose tissue, and TGF-\g=b\mRNA is expressed in both proliferating and differentiating 3T3-L1 preadipocytes. Together with the data presented here, this would indicate that these peptides have a role in adipocyte development by an autocrine or paracrine mechanism although the source of PDGF in vivo is at present unknown. Journal of Endocrinology (1991) 131, 203\p=n-\209
INTRODUCTION
& Ailhaud, 1987) and proliferation of 3T3-L1 cells (Smith, Wise, Berkowitz et al. 1988). IGF-I mRNA is also expressed in differentiated Ob 1771 preadipocytes, in murine adipose tissue stromal-vascular cells (Doglio, Dani, Fredrikson et al. 1987) and in differentiated primary cultures of porcine adipocyte precursors (Gaskins, Kim, Wright et al. 1990). Transforming growth factor-ß (TGF-ß) also regu¬ lates differentiation of a number of cell types, in¬ cluding adipocytes (Ignotz & Massague, 1985; Torti, Torti, Larrick & Ringold, 1989), but the proliferative response to TGF-ß is dependent on the cell type and can be inhibitory or stimulatory (Moses, Tucker, Leof et al. 1985). Stimulation of DNA synthesis may be an indirect effect of TGF-ß via the production of platelet-
IGF-I at certain concentrations. Stimulation of proliferation of some cell types by TGF-\g=b\has been linked to the secondary production
development of preadipocyte cell lines and cul¬ primary adipocyte precursors has aided the greatly study of adipocyte development and The
ture methods for
has enabled details of the factors which specifically regulate proliferation and differentiation to emerge. Most studies have concentrated on the regulation of differentiation in specific cell lines such as 3T3-L1 and Ob 1771 and less is known about factors which regulate proliferation in primary cultures of adipocyte precursor cells. It has been shown that insulin-like growth factor-I (IGF-I) can stimulate proliferation and differen¬ tiation of rat adipocyte precursor cells (Deslex, Negrel
derived growth factor (PDGF)-like proteins coded by the c-sis proto-oncogene (Leof, Proper, Goustin et al. 1986; Soma & Grotendorst, 1989) which then stimu¬ late proliferation. TGF-ß mRNA is also expressed in proliferating and differentiating 3T3-L1 preadipocytes (Weiner, Shah, Smith et al. 1989). This evidence is consistent with the idea that these growth factors may have an autocrine/paracrine role in adipocyte development but most studies have investigated the role of growth factors added singly to cells in culture, whereas cells in vivo are exposed to a combination of factors. This study was designed to investigate the relationship between IGF peptides, PDGF and TGF-ßl on the regulation of DNA syn¬ thesis in primary cultures of adipocyte precursor cells. MATERIALS AND METHODS
TGF-ßl was obtained from British Biotechnology, Cowley, Oxon, U.K. or was a gift from Dr R. Derynk (Genentech, South San Francisco, CA, U.S.A.). Both batches were equipotent. Recombinant cDNA-
derived human IGF-I, IGF-II and PDGF-AB were from Bachern, Saffron Waiden, Essex, U.K. Bovine serum albumin (BSA, radioimmunoassay grade) was from Sigma, Poole, Dorset, U.K. Dulbecco's A phosphate-buffered saline (PBS) was prepared from tablets according to the manufacturer's instructions
(Oxoid Co., Basingstoke, Hants, U.K.). Trypsin (2-5%, w/v), gentamicin and Ultroser were from Gibco-BRL, Paisley, Strathclyde, U.K. Tissue culture plates (Costar; 24-multiwell plates) and medium 199 (Ml99) were from Northumbria Biologicals, Cramlington, Northumberland, U.K. 6-[3H] Thymidine (25 Ci/mmol) was from Amersham Inter¬ national pic, Amersham, Bucks, U.K. Animals
Female broiler chickens of a commercial line were obtained from D. B. Marshall, Newbridge, Midlothian, U.K. They were housed in brooders with a lighting pattern of 14 h light : 10 h darkness and fed on stan¬ dard broiler diets. Birds were killed at 2-3 weeks of age by cervical dislocation, and abdominal adipose tissue was removed for the preparation of adipocyte
precursor cells.
Cell culture
Adipocyte precursor cells were prepared and cultured in supplemented M199 according to the method of Butterwith & Griffin (1989). They were plated in multiwell plates at a density of 32 000 cells/well
(2 cm2) in 1ml
M199 containing 4% (v/v) Ultroser and incubated for 66 h at 37 °C in a humidified atmos¬ phere containing 5% C02. They were washed and incubated in M199 for a further 24 h before addition of growth factor together with [3H]thymidine (0-25 pCi/well). After incubation for 18 h the super¬ natant was aspirated and the cell monolayer washed with PBS (1 ml) and treated with trypsin (400 pi; 1%, w/v) and NaOH (100 pi; 5 mol/1). The cell suspension was transferred to a microfuge tube, 100 pi BSA (1 %, w/v) and 350 pi perchloric acid (2 mol/1) were added and the samples incubated overnight at 4 °C. After centrifugation at 2800 # for 10 min, the precipitate was washed with ice-cold perchloric acid (1 mol/1) before being resuspended in 300 pi perchloric acid (1 mol/1) and hydrolysis at 70 °C for 30 min. After further centrifugation, a sample of the supernatant was counted for radioactivity in Optiphase X (Beckman
Instruments, High Wycombe, Bucks, U.K.) using
a
Wallac 1410 scintillation counter (Pharmacia LKB Biotechnology, Milton Keynes, Bucks, U.K.). Initial experiments demonstrated that there was a good correlation between [3H]thymidine incorporation and cell number. In order to investigate the time required to observe effects of TGF-ßl, cells were preincubated with TGFßl for 3 h. The medium was removed and cells were washed with fresh medium (1 ml) and then incubated for 18 h in M199 containing [3H]thymidine and IGF-I. Control incubations received no TGF-ßl during preincubation, which was followed by the addition of IGF-I, TGF-ßl, both or medium alone. The pre¬ incubation period was also extended to 4 and 5 h. In a further experiment, the effect of medium con¬ ditioned by cells previously exposed to TGF-ßl was examined. Cells were preincubated with TGF-ßl (1 pg/1) for 6h, washed with fresh medium (1ml) and then incubated for 18 h in M199 alone. This conditioned medium was added to a different batch of growth-restricted cells and [3H]thymidine incor¬ poration into DNA was measured as previously described. Medium conditioned by cells which had not been exposed to TGF-ßl was used as a control and the effect of IGF-I (5 pg/1) alone and a combi¬ nation of IGF-I and TGF-ßl were used as positive controls in each case. Statistics
The ALLFIT program (De Lean, Munson & Rodbard, 1978) was used to analyse dose-response curves and calculate half-maximum effective concentrations for thymidine incorporation (ED50). A two-way analy¬ sis of variance was used to test for synergistic inter¬ actions using the MINITAB statistical package on an Apple Macintosh SE/30 microcomputer. Other data
were
analysed by Student's
statistical
/-test
package.
the
using
1200
same
()
1000
RESULTS
800
The cells were identified as adipocyte precursors on the basis of their ability to differentiate in culture on reaching confluence. Differentiation was character¬ ized by an increase in lipoprotein lipase activity and lipogenesis (Butterwith & Griffin, 1989) and by the accumulation of lipid droplets.
600
W
*'
400 J
o.
1200
•
1 IGF-1
(b)
synthesis incorporation of [3H]thymidine into DNA by adipocyte precursor cells in the absence of added
800
(u.g/1)
.»
The
600
Ultroser or exogenous growth factor varied between experiments but was in the range 500-1800 c.p.m./well with a mean value of 892 + 511 (s.d.; 6) for the addition of M199 here. The experiments reported
8
400
=
containing 4% Ultroser increased the [3H]thymidine incorporation to 75 129+19 789 c.p.m./well (n 5),
4000
1 IGF-II
0
(c)
=
an
84-fold stimulation.
Effect of IGF-I, IGF-II and
3000
The effects of IGF-I, IGF-II and TGF-ßl on [3H] thymidine incorporation into DNA by adipocyte pre¬ cursors when added alone are shown in Fig. 1. Both IGF-I and IGF-II stimulated [3H]thymidine incorpor¬ ation in a dose-dependent manner, with a maximum response (about twofold) at 5 pg/1 with an ED50 of 2-06+1-47 and 0-92 + 0-75 (s.e.m.; d.f. 8 and 7) pg/1 for IGF-I and IGF-II respectively. The ED50 values were not significantly different as determined by a =
P
=
0-54
on
16d.f.)
TGF-ßl also stimulated [3H]thymidine incorpor¬ ation by about threefold with the maximum at con¬ centrations of 1 pg/1 and an ED50 of 0-34+ 0025 (s.e.m.; d.f. 10) pg/1. The response to TGF-ßl was more sensitive than the IGF peptides when the ED50 values were compared (F=4-89, P
