- Email: [email protected]
Dacron stimulation of macrophage transforming growth factor-β release
EJ.SEWFiR SCiENCEc
Dacrsn stimulation of maw -____ grwvth factor+3release
----.-.._-----_.
H. P. Creisler, D. Petsikas, D. J. Cziperle, P. M. Murchan. S. C. Henderson* and T. M. Lam ioyola University Medical Center. Dtpartmen t of Surgery. Maywood. ///ino& and *Hines VAN0spiY-d. Department of Surgery, Hines, fflinois, USA
This study evaluated the effect of Dacron on the release of macrophage transforming growth factor-p (TGF-P). an endothelial cell growth inhibitor. Rabbit peritoneal macrop were grown in minimum essential medium (MEM) with 10% fetal bovine serum (PBS) in the presence or absence of Dacron (0.5 mm x 3 mm particles). Media were collected three times each week for 7 weeks. For the TGF-p bioassay, mink lung epithelial cells (CCL64)were grown in MEM with 10% FBS.Test-conditioned media, 100 PI. were added (n = 4), and incubated 4Bh. 3H-Thymidine (3H-TdR) uptake was determined and compared with 3H-TdR uptake using known pure TGF-P standards. Media samples were additionally pre-Incubated wKh a neutralizing anti-TGF-/3, antibody and the 3H-TdR uptake again quantiited. TGF-fi activity in the conditioned media of macrophages exposed to Dacron exceeded the control media groups in all weeks, reaching significance (f
events occurring after graft implantation IS an extensive chronic inflammatory response with deposition of monocyte-derived macrophages interacting with the biomaterial. It was hypothesized that biomaterials may differentially activate macrophages, resulting in the production and secretion of different bioactive substances including growth stimulatory and inhibitory molecules. In the present study, it was further postulated that a mechanism by which Dacron may inhibit endothelialization and perhaps stimulate anastomotic pseudointimal hyperplasia may be its interaction with the macrophage, resulting in macrophage production of transforming growth factor-beta (TGF-B:.
Materials and methods Macrophage culture
Correspondence to: Dr H. P. Greisler, Loyola University Medical Center, Department of Surgery, 2160 South First Avenue, Maywood, Illinois 6015.3, USA
CARDlOVA!X!ULAR SURGERY APRIL 1996 VOL 4 NO 2
Peritoneal macrophages were harvested from adult New Zealand White rabbits by peritoneal lavage with’Hanks balanced salt solution (HBSS) with calcium and
169
Dacron stimulattbn of TGF$ release: H, P Grekler et al.
magnesium and with lOU/ml heparin. Collected cells were centrifuged at 900g for 10min and washed three times with phosphate-buffered saline. Cells from five rabbits were pooled and suspended in minimum essential medium (MEM) containing 10% equine platelet-poor serum, 100 U/ml penicillin, 0.1 mg/ml streptomycin, L-glutamine 200 mmol/l, Hepes buffer, sodium pyruvate 100 mmol/l, and non-essential amino acids 10 mmol/l. Collected cells, 1 X lo6 cells/ml, were seeded into three T-25 Falcon (Becton-Dickinson Labware, Lincoln Park, New Jersey, USA) tissue culture flasks containing 3 ml MEM. Cell viability and cell counts were determined using Trypan blue dye exclusion with a hemocytometer. Following a 2-h adherence separation, adherent cells were washed with MEM and fed with 5 ml complete media. Ambient conditions were humidified with 5% COa. Macrophage culture purity was determined by phase contrast microscopy, non-specific esterase staining by the method of Yam et ~1.~, identification of Fc (immunoglobulin G) receptors on the cell membrane by the method of Bianco and Pitkowski4, and immunoperoxidase using the rabbit macrophage-specific anti-RAM 11 antibody. Immunoperoxidase staining was repeated on cells sampled at the conclusion of the experiments to verify continued culture purity. Finely shredded Dacron, 0.5 mm x 3 mm particles, 1.6 mg/cm2, was added to the culture media of one-half of the cells. Flasks were routinely fed with complete media three times weekly and conditioned media collected at the times of refeeding. Collected media were centrifuged at 3000 rpm for 15 min and supernatants filtered through a 0.22-u Millipore filter (Millipore Corporation Continental Water Systems, Bedford, Massachusetts, USA). The three media collections each week were pooled and frozen to -70°C for use in the bioassays. Mink lung epithelial cell culture and bioassay CCL64 mink lung epithelial cells (ATCC, Rockville, Maryland, USA) are exquisitely sensitive to the growth inhibitory activity of TGF-B. CCL64 cells were cultured in MEM with non-essential amino acids, 10% fetal bovine serum (FBS), penicillin lOOU/mg, and streptomycin 0.1 mg/ml. These cells were trypsinized 3 days after their last passage, seeded into 24-well plates (5000 cells/well) and incubated at 37°C in 5% humidified CO2 for 1 h to permit attachment. Cells were then plated in complete growth media, 100 ul of thawed conditioned media added, and the cells allowed to incubate for 48 h. CCL64 cells used for positive controls received 100 ul (0.5 ng) pure TGF-Bi (R & D Systems, Minneapolis, Minnesota, USA); negative controls received 100 ul of phosphate-buffered saline (PBS). Each condition (each weekly media collection of both control and Dacron groups), was repeated in quadruylicate. Following 48-h incubation, tritiated thymidine ( H-TdR), 1 uCi (10 ~1)
170
was added to each well and allowed to incubate for 16 h at 37°C. Wells were then processed by scintillation counting to monitor 3H-TdR incorporation into newly synthesized DNA. Cells were washed with physiological saline and fixed twice with 500 ul absolute methanol at 4°C for 5 min each. Cells were then washed with distilled H20 and incubated with 5% trichloroacetic acid (TCA) for 10min each. Cells were re-washed three times with distilled H20 and TCA-precipitable material resolubilized in 0.3 M NaOH, 300 ~1, and incubated for 20 min. Samples were then transferred to lo-ml Ready Protein scintillation cocktail for counting in a Beckman LS6800 scintillation counter. TGF-l3 activity in samples was determined based upon the standard curve generated by serial dilutions of known quantities of pure TGF-l3 (Figure 1) and expressed as percent of maximal TGF-l3 and percent inhibitory activity. Percent inhibition TGF-B activity were calculated by the following formulae: X
100
Positive control % TGF-l3 activity
% Inhibition
=
Maximal
TGF-l3 activity
To confirm the presence of TGF-B, the conditioned media groups from the macrophages exposed to Dacron were additionally pre-incubated with 200 @ml of a specific neutralizing anti-TGF-B1 antibody (R & D Systems, Minneapolis, Minnesota, USA) to selectively inhibit TGF-Bi activity. This concentration of antibody was determined by preliminary studies to identify the neutralizing doses0 defined as that concentration of antibody required to yield one-half maximal inhibition
too
00 00
I
f--
I
5 8” 6 P aOmto o.&mo,
0.ooo1
0.001
0.01
0.1
TGF-8 (rigs/well) Figure 1 TGF-P standard curve. CCL64 cell proliferation on the Y axis is plotted against serial dilutions of TCiF& on the X axis. Values are mean (s.d.)
CARDIOVASCULAR SURGERY APRIL 1996 VOL 4 NO 2
Dacron stirnutation of TGF-,8release: ki. P. GrwHer et al.
of TGF-PI when the TGF& is present at five times its effective doses*. This was determined by a CCL64 inhibition curve generated in response to serial dilutions of pure TGF-PI. These pre-incubated media samples were then added to additional wells of CCL64 cells and 3H-TdR uptake again quantitated. Statistical analysis Data expressed as percent of pure TGF-/3 activity were used to calculate mean values of replicate wells; mean values were then compared as a function of culture duration and as a function of the presence or absenceof Dacron using two-way analysis of variance and independent Student’s t tests. Comparison between groups with and without the pre-incubation with anti-TGF-P antibody were made using paired Student’s t tests.
Results Macrophages cultured either in the presence or absence of Dacron particles appeared morphologically identical, with no demonstrable change in cell numbers as a function of either time or presence of the Dacron particles. Macrophage culture purity in both groups was > 95%. Macrophages cultured in the presence of Dacron were seen to adhere to the Dacron particles (Figure 2). No intracytoplasmic inclusions were detected, suggesting a lack of phagocytosis. TGF-fi activity in conditioned media of macrophages exposed to Dacron was statistically significantly greater (P < 0.05) than in control macrophage-conditioned media in samples collected during weeks 3,4,5,6 and 7 of culture (Table 1; Figure 3). In the first 2 weeks of
Tab&e 1 Dacron-induced TGF-0 secretion in rabbit pe&oneal macrophages. Data are expressed as TGF-(3 activity in media compared with % of pure TGF-6 standard (5 q/ml) Week
Control
Dacron
Dacron -L anti-TGF-fil antibody -.---.
1 2 3 4
21 BO(8.08) 22.70(5.54) 23.23(5.83) 5.57(7.24) 4.60(6.11) 12.3-0(16.26) 14.22( 13.75)
28.57( 1.68) 34.52(9.57) 35.67(4.91)* 1790(4.71)* 36.37(3.01)* 37.60(9.17)” 42.80(8.78)*
530( 10.6oi ’ 17.95(7.67) 7.79t6.95; + 24.46(5.25) 430(8.60) ( 25.26{19.221 14.66(6.65) +
: 7
Values are mean(s.d.). *P< 0.05 Venus controi medra: ‘P
culture these differences did not reach statistical significance. The change in TGF-P activity as a function of time showed a trend towards increasing levels in the Dacron group, which contrasted with a trend towards decreasing TGF-fi activity levels in the control group. Pre-incubation of the conditioned media with neutralizing anti-TGF& antibody significantly blocked TGFp activity in the conditioned media of macrophages exposed to Dacron collected during weeks 1,2,3,5 and 7, with differences in weeks 4 and 6 not achieving statistical significance.
Discussion Data obtained from the present study demonstrate increased TGF-PI protein release by macrophages cultured in vitro in the presence of Dacron particles as compared with similar macrophages cultured in the absence of Dacron. The TGF-B activitv documented in nf:dia that was not neutrahzed bv’the anri-TGF-PI
i
2
3
CULTURE Figure 2 Phase contrast photomicrograph showing rabbit peritoneal macrophages following 6 weeks in culture in the presence of Dacron particles. Cells can be seen adhering both to the polystyrene culture flask and to the Dacron particles but without intracytoplasmic inclusions. (Original magnification x ZOO)
CARDIOVASCULAR SURGERY APRIL 1906 VOL 4 NO 2
4
DURATION
(WEEKS)
Figure 3 Percent of maximal T&-6 act&& in conditioned media of macrophages with or without Dacron exposure and with or without pre-incubation with an anti-TGF-0, ant&c&, the media being collected in successive weeks of culture. ‘Pro.05 Dacron *M Dacron plus anti-TGF-6, antibody: ‘PC 0.05 Dacron VEXKJScontroi
171
Dacronstimulation of TGF-/3release:H. P. Gmslef et al. antibody probably represents the presence of other isoforms of TGF-j3 not blocked by the specific anti-TGF-Pi antibody utilized. All -studies were performed with the same anti-TGF-Pi antibody which is isoform-specific. The previously documented inhibitory activity of Dacron on endothelialization may be explained in part by this Dacron-induced release of TGF-l3 from macrophages. The present study does not differentiate among a Dacron-induced up-regulation of TGF$ mRNA transcription, TGF-l3 protein synthesis, and/or release of stored intracytoplasmic TGF-P protein. Three TGF-P isoforms have been isolated, with TGF-Pi the most prevalent and well characterized. TGF-Pi is a 25-kDa homodimeric peptide secreted in a latent form which requires proteolysis or acidification before receptor bindingj. Three TGF-I3 receptors have also been identified6. It is not possible from the present study to quantitate secretion of latent TGF-P inasmuch as this assay specifically identifies active TGF-l3. Activation of TGF-l3 generally results from proteolysis in an acidic environment. The mitogenic activity of TGF-l3 on both endothelial cells and smooth muscle cells is complex. Endothelial cell growth in vitro is inhibited by TGF-/3’, endothelial ase is delayed’ and c-myc cell entry into the S expression is decreasedgh. However, endothelial cells cultured in three-dimensional collagen gels are not TGF-l3 inhibited” and TGF-/3 installation into newborn and chorioallantoic membranesi induces mice’l angiogenesis. Higher concentrations of TGF-P however resulted in an inhibition of angiogenesis in these models13. TGF-I3 probably plays a significant role in capillary angiogenesis. Exposure of bovine aortic or capillary endothelial cells to fibroblast growth factor induces a significant increase in plasminogen activator activity, probably essential for basement membrane degradation required to permit endothelial sprout migration in the early phase of new capillary formation. However, high concentrations of TGF-Pi (5-10 rig/ml) cause an inhibition of TGF-B-induced angiogenesis by bovine microvascular endothelial cells cultured on three-dimensional collagen or fibrin gels i4 . Introduction of chondrocyteconditioned media into cultures of endothelial cell monolayers induces increased plasminogen activator inhibitor type 1 (PAI-1) protein and mRNA production, a response blocked by pre-incubation of the media with a neutralizing anti-TGF-l3 antibody”. Chondrocyte/ endothelial cell co-culture significantly inhibits capillary tube formation, this activity also being blocked by anti-TGF-I3 antibodies l6 . These data suggest that the anti-angiogenic properties of cartilage may be mediated by chondrocyte-derived TGF-6 induction of PAIproduction by endothelial cells”. The data collected in the experiment showing TGF-l3 secretion by macrophages exposed to Dacron in vitro should be further investigated by studies to document similar effects in
172
vivo and to determine the relationship of secreted TGF-(3 to endothelialization of Dacron prostheses in vivo. The effect of TGF-P on fibroblast and smooth muscle cell growth shows a bimodal concentration effect. Low concentrations of TGF-I3 stimulate fibroblast and smooth muscle cell proliferation whereas higher concentrations inhibit it. This is probably due to the ability of low TGF-P concentrations to up-regulate plateletderived growth factor (PDGF) transcription in fibroblasts and smooth muscle cells whereas higher concentrations simultaneousill down-regulate PDGF-I3 receptor subunit expression . TGF-0 is also a potent stimulator of extracellular matrix molecular synthesis. Subcutaneous TGF-l3 injection induces collagen synthesis’* and TGF-I3 has been shown to induce gene expression and protein synthesis of collagen types I and III and fibronectin’9-21. Spontaneous endothelialization of Dacron grafts implanted clinically has not been reported. The pannus ingrowth of endothelial cells across the anastomosis is limited to approximately 1 cm in length and this perianastomotic region is characterized by exuberant smooth muscle cell and fibroblast proliferation and collagen deposition. These histologic responses result in a relatively thrombogenic acellular flow surface over the majority of the graft’s length and the development of peri-anastomotic pseudointimal hyperplasia. All of these phenomena may in part be mediated by TGF-P production which could induce both smooth muscle cell and fibroblast proliferation and collagen synthesis as well as inhibit endothelial cell ingrowth. However, extrapolation of the in vitro data presented here directly to the in vivo scenario is not possible. Additional studies are warranted to investigate TGF-P protein and message levels in vivo as a function of graft type, anatomic location vis-a-vis the peri-anastomotic region, time following implantation and cell of origin. These studies are currently in progress using immunohistochemical and in situ hybridization techniques.
Acknowledgements These studies were supported by a grant from the National Institutes of Health, ROl HL 41272. The authors also thank Dr A. Gown for the gift of rabbit macrophage-specific anti-RAM 11 antibody.
References 1.
2. 3.
Foxall TL, Auger KR, Callow AD et al. Adult human endothelial cell coverage of small-caliber Dacron and polytetrafluoroethylene vascular prostheses in r&o. J Surg Res 1986; 41: 158-72. Greisler HP, Schwartz TH, Ellinger J et al. Dacron inhibition of arterial regenerative activity. ] Vusc Surg 1986; 3: 747-56. Yam LY, Li CY, Crosby WH. Cytochemical identification of monocytes and granulocytes. Am J Clin Pathol 1971; 55: 283-90.
CARDIOVASCULAR SURGERY APRIL 1996 VOL 4 NO 2
Dacron stimulation of TGF$ rele%w h’. I? Greisier et al. Bianco C, Pitowski B. Fc and C3 receptors. In: Adams DO, Edelson PJ, Koren HS, eds. Methods for Studying Mononuclear Phugocytes. New York: Academic Press, 1981: 273-328. 5. Assoian RK, Komoriya A, Meyers CA et al. Transforming growth factor-beta in human platelets. Identification of a major storage site, purification, and characterization. / Biol Chem 1983; 258: 7155-60. 6. Barnard JA, Lyons RM, Moses HL. The ceil biology of transforming growth factor f5. Biochim Biophys Actu 1990; 1032: 79-87. 7. Frater-Schroder M, Risau W, Hailmann R et al. Tumor necrosis factor type (Y, a potent inhibitor of endotheiiai ceil growth in vitro, is angiogenic in viuo. Proc Nat1 Acad Sci USA 1987; 84: 5277-81 RL, Twardzik DR, Schwartz SM. Inhibition of 8. Heimark endotheiial regeneration by type-beta transforming growth factor from platelets. Science 1986; 233: 1078-80. 9. Takehara K, LeRoy EC, Grotendorst GR. TGF-B inhibition of endotheiial ceil proliferation: alteration of EGF binding and EGF-induced growth regulatory (competence) gene expression. Cdl 1987; 49: 415-22. 10. Madri JA, Pratt BM, Tucker AM. Phenotypic modulation of rndotheiial ceils by transforming growth factor-beta depends upon the composition and organization of the extraceiiuiar matrix. I Cell Biol 1988; 106: 1375-84. 11. Roberts AB, Spom MB, Assoian RK et al. Transforming growth factor type B: rapid induction of fibrosis and angiogenesis in viva and stimulation of collagen formation in vitro. Proc Nat1 Acad Sci USA 1986; 83: 4167-71. growth factor beta 12. Yang EY, Moses HL. Transforming l-induced changes in cell migration, proliferation, and angiogenesis in the chicken chorioailantoic membrane. ] Cell Biol1990; 111: 731-41. 13. Moses HL, Yang EY, Pietenpol JA. TGF-P stimulation and inhibition of ceil proliferation: new mechanistic insights. Cell 4.
CAfEliOVASCULAA SURGERY APRIL 1SQ6 VOL 4 NO 2
14. 15.
16.
17.
18.
19. 20.
21.
3 990; 63: 245-7. Pepper MJ, Vassaih JD, Orci I. ef a/. Bipbasic rffect of transforming growth factor-B, on in ~itn~ air~iogenesn,. F:xp Crll Res 1993; 204: 356-63. Pepper MS, Montesano R, Orci L et al. Plasmrnogen activatorinhibitor-l is induced in microvascular endotheiial cells by a chondrocyte-derived transforming growrh tactor-beta. Hiochen? Biophys Res Comwnm 1991; 176: 633-8. Pepper MS, Montesano R, Vassalii JD ri c?i (bondrocytrs inhibit endothelial sprout formation m t~itro: evidence for the involvement of a transforming growth factclr-beta. 1 Ckl! Phs?/tol 1991; 146: 170-9. Leof EB, Proper JA, Goustin AS et al. induction ot c-srs mRNA and activity similar to platelet-derived growth [actor by transforming growth factor beta: a proposed model for mdirect mitogenesis involving autocrine activity. i~rr)~; Yzzt! Arad 5’~ USA 1986; 83: 2453-7. Roberts AB, Sporn MB, Assoian RK er al. Transforming growth factor type B: rapid induction of fibrosis and dngiogenesis in viva and stimulation of collagen formation in VI~YO.Pro<. Nat/ Acad Sci USA 1986; 83: 4167-71. ignotz RA, Endo T, Massague J. Regulation of ftbronectin and type I collagen mRNA levels by transforming growth factor-(-l. / Biol Chem 1987; 262: 6443-6. Varga J, Rosenbloom J, Jimenez SA. Transforming growth factor-p (TGF-R) causes a persistent increase in steady-state amounts of type I and type II1 collagen and iihrtmectm mRNAs in normal dermai fibrobiasts. Biochenrrstn* 1987: 247. 597-604. Rossi P, Karsentry G, Roberts AB et JI. A nuclaar tactor I binding site mediates the transcriptianal activation of a type I collagen promoter by transforming growth tnctor-/S. (‘elf 1988; 52: 405-14.
Paper accepted 11 July 1994
173
Dacrsn stimulation of maw -____ grwvth factor+3release
----.-.._-----_.
H. P. Creisler, D. Petsikas, D. J. Cziperle, P. M. Murchan. S. C. Henderson* and T. M. Lam ioyola University Medical Center. Dtpartmen t of Surgery. Maywood. ///ino& and *Hines VAN0spiY-d. Department of Surgery, Hines, fflinois, USA
This study evaluated the effect of Dacron on the release of macrophage transforming growth factor-p (TGF-P). an endothelial cell growth inhibitor. Rabbit peritoneal macrop were grown in minimum essential medium (MEM) with 10% fetal bovine serum (PBS) in the presence or absence of Dacron (0.5 mm x 3 mm particles). Media were collected three times each week for 7 weeks. For the TGF-p bioassay, mink lung epithelial cells (CCL64)were grown in MEM with 10% FBS.Test-conditioned media, 100 PI. were added (n = 4), and incubated 4Bh. 3H-Thymidine (3H-TdR) uptake was determined and compared with 3H-TdR uptake using known pure TGF-P standards. Media samples were additionally pre-Incubated wKh a neutralizing anti-TGF-/3, antibody and the 3H-TdR uptake again quantiited. TGF-fi activity in the conditioned media of macrophages exposed to Dacron exceeded the control media groups in all weeks, reaching significance (f
events occurring after graft implantation IS an extensive chronic inflammatory response with deposition of monocyte-derived macrophages interacting with the biomaterial. It was hypothesized that biomaterials may differentially activate macrophages, resulting in the production and secretion of different bioactive substances including growth stimulatory and inhibitory molecules. In the present study, it was further postulated that a mechanism by which Dacron may inhibit endothelialization and perhaps stimulate anastomotic pseudointimal hyperplasia may be its interaction with the macrophage, resulting in macrophage production of transforming growth factor-beta (TGF-B:.
Materials and methods Macrophage culture
Correspondence to: Dr H. P. Greisler, Loyola University Medical Center, Department of Surgery, 2160 South First Avenue, Maywood, Illinois 6015.3, USA
CARDlOVA!X!ULAR SURGERY APRIL 1996 VOL 4 NO 2
Peritoneal macrophages were harvested from adult New Zealand White rabbits by peritoneal lavage with’Hanks balanced salt solution (HBSS) with calcium and
169
Dacron stimulattbn of TGF$ release: H, P Grekler et al.
magnesium and with lOU/ml heparin. Collected cells were centrifuged at 900g for 10min and washed three times with phosphate-buffered saline. Cells from five rabbits were pooled and suspended in minimum essential medium (MEM) containing 10% equine platelet-poor serum, 100 U/ml penicillin, 0.1 mg/ml streptomycin, L-glutamine 200 mmol/l, Hepes buffer, sodium pyruvate 100 mmol/l, and non-essential amino acids 10 mmol/l. Collected cells, 1 X lo6 cells/ml, were seeded into three T-25 Falcon (Becton-Dickinson Labware, Lincoln Park, New Jersey, USA) tissue culture flasks containing 3 ml MEM. Cell viability and cell counts were determined using Trypan blue dye exclusion with a hemocytometer. Following a 2-h adherence separation, adherent cells were washed with MEM and fed with 5 ml complete media. Ambient conditions were humidified with 5% COa. Macrophage culture purity was determined by phase contrast microscopy, non-specific esterase staining by the method of Yam et ~1.~, identification of Fc (immunoglobulin G) receptors on the cell membrane by the method of Bianco and Pitkowski4, and immunoperoxidase using the rabbit macrophage-specific anti-RAM 11 antibody. Immunoperoxidase staining was repeated on cells sampled at the conclusion of the experiments to verify continued culture purity. Finely shredded Dacron, 0.5 mm x 3 mm particles, 1.6 mg/cm2, was added to the culture media of one-half of the cells. Flasks were routinely fed with complete media three times weekly and conditioned media collected at the times of refeeding. Collected media were centrifuged at 3000 rpm for 15 min and supernatants filtered through a 0.22-u Millipore filter (Millipore Corporation Continental Water Systems, Bedford, Massachusetts, USA). The three media collections each week were pooled and frozen to -70°C for use in the bioassays. Mink lung epithelial cell culture and bioassay CCL64 mink lung epithelial cells (ATCC, Rockville, Maryland, USA) are exquisitely sensitive to the growth inhibitory activity of TGF-B. CCL64 cells were cultured in MEM with non-essential amino acids, 10% fetal bovine serum (FBS), penicillin lOOU/mg, and streptomycin 0.1 mg/ml. These cells were trypsinized 3 days after their last passage, seeded into 24-well plates (5000 cells/well) and incubated at 37°C in 5% humidified CO2 for 1 h to permit attachment. Cells were then plated in complete growth media, 100 ul of thawed conditioned media added, and the cells allowed to incubate for 48 h. CCL64 cells used for positive controls received 100 ul (0.5 ng) pure TGF-Bi (R & D Systems, Minneapolis, Minnesota, USA); negative controls received 100 ul of phosphate-buffered saline (PBS). Each condition (each weekly media collection of both control and Dacron groups), was repeated in quadruylicate. Following 48-h incubation, tritiated thymidine ( H-TdR), 1 uCi (10 ~1)
170
was added to each well and allowed to incubate for 16 h at 37°C. Wells were then processed by scintillation counting to monitor 3H-TdR incorporation into newly synthesized DNA. Cells were washed with physiological saline and fixed twice with 500 ul absolute methanol at 4°C for 5 min each. Cells were then washed with distilled H20 and incubated with 5% trichloroacetic acid (TCA) for 10min each. Cells were re-washed three times with distilled H20 and TCA-precipitable material resolubilized in 0.3 M NaOH, 300 ~1, and incubated for 20 min. Samples were then transferred to lo-ml Ready Protein scintillation cocktail for counting in a Beckman LS6800 scintillation counter. TGF-l3 activity in samples was determined based upon the standard curve generated by serial dilutions of known quantities of pure TGF-l3 (Figure 1) and expressed as percent of maximal TGF-l3 and percent inhibitory activity. Percent inhibition TGF-B activity were calculated by the following formulae: X
100
Positive control % TGF-l3 activity
% Inhibition
=
Maximal
TGF-l3 activity
To confirm the presence of TGF-B, the conditioned media groups from the macrophages exposed to Dacron were additionally pre-incubated with 200 @ml of a specific neutralizing anti-TGF-B1 antibody (R & D Systems, Minneapolis, Minnesota, USA) to selectively inhibit TGF-Bi activity. This concentration of antibody was determined by preliminary studies to identify the neutralizing doses0 defined as that concentration of antibody required to yield one-half maximal inhibition
too
00 00
I
f--
I
5 8” 6 P aOmto o.&mo,
0.ooo1
0.001
0.01
0.1
TGF-8 (rigs/well) Figure 1 TGF-P standard curve. CCL64 cell proliferation on the Y axis is plotted against serial dilutions of TCiF& on the X axis. Values are mean (s.d.)
CARDIOVASCULAR SURGERY APRIL 1996 VOL 4 NO 2
Dacron stirnutation of TGF-,8release: ki. P. GrwHer et al.
of TGF-PI when the TGF& is present at five times its effective doses*. This was determined by a CCL64 inhibition curve generated in response to serial dilutions of pure TGF-PI. These pre-incubated media samples were then added to additional wells of CCL64 cells and 3H-TdR uptake again quantitated. Statistical analysis Data expressed as percent of pure TGF-/3 activity were used to calculate mean values of replicate wells; mean values were then compared as a function of culture duration and as a function of the presence or absenceof Dacron using two-way analysis of variance and independent Student’s t tests. Comparison between groups with and without the pre-incubation with anti-TGF-P antibody were made using paired Student’s t tests.
Results Macrophages cultured either in the presence or absence of Dacron particles appeared morphologically identical, with no demonstrable change in cell numbers as a function of either time or presence of the Dacron particles. Macrophage culture purity in both groups was > 95%. Macrophages cultured in the presence of Dacron were seen to adhere to the Dacron particles (Figure 2). No intracytoplasmic inclusions were detected, suggesting a lack of phagocytosis. TGF-fi activity in conditioned media of macrophages exposed to Dacron was statistically significantly greater (P < 0.05) than in control macrophage-conditioned media in samples collected during weeks 3,4,5,6 and 7 of culture (Table 1; Figure 3). In the first 2 weeks of
Tab&e 1 Dacron-induced TGF-0 secretion in rabbit pe&oneal macrophages. Data are expressed as TGF-(3 activity in media compared with % of pure TGF-6 standard (5 q/ml) Week
Control
Dacron
Dacron -L anti-TGF-fil antibody -.---.
1 2 3 4
21 BO(8.08) 22.70(5.54) 23.23(5.83) 5.57(7.24) 4.60(6.11) 12.3-0(16.26) 14.22( 13.75)
28.57( 1.68) 34.52(9.57) 35.67(4.91)* 1790(4.71)* 36.37(3.01)* 37.60(9.17)” 42.80(8.78)*
530( 10.6oi ’ 17.95(7.67) 7.79t6.95; + 24.46(5.25) 430(8.60) ( 25.26{19.221 14.66(6.65) +
: 7
Values are mean(s.d.). *P< 0.05 Venus controi medra: ‘P
culture these differences did not reach statistical significance. The change in TGF-P activity as a function of time showed a trend towards increasing levels in the Dacron group, which contrasted with a trend towards decreasing TGF-fi activity levels in the control group. Pre-incubation of the conditioned media with neutralizing anti-TGF& antibody significantly blocked TGFp activity in the conditioned media of macrophages exposed to Dacron collected during weeks 1,2,3,5 and 7, with differences in weeks 4 and 6 not achieving statistical significance.
Discussion Data obtained from the present study demonstrate increased TGF-PI protein release by macrophages cultured in vitro in the presence of Dacron particles as compared with similar macrophages cultured in the absence of Dacron. The TGF-B activitv documented in nf:dia that was not neutrahzed bv’the anri-TGF-PI
i
2
3
CULTURE Figure 2 Phase contrast photomicrograph showing rabbit peritoneal macrophages following 6 weeks in culture in the presence of Dacron particles. Cells can be seen adhering both to the polystyrene culture flask and to the Dacron particles but without intracytoplasmic inclusions. (Original magnification x ZOO)
CARDIOVASCULAR SURGERY APRIL 1906 VOL 4 NO 2
4
DURATION
(WEEKS)
Figure 3 Percent of maximal T&-6 act&& in conditioned media of macrophages with or without Dacron exposure and with or without pre-incubation with an anti-TGF-0, ant&c&, the media being collected in successive weeks of culture. ‘Pro.05 Dacron *M Dacron plus anti-TGF-6, antibody: ‘PC 0.05 Dacron VEXKJScontroi
171
Dacronstimulation of TGF-/3release:H. P. Gmslef et al. antibody probably represents the presence of other isoforms of TGF-j3 not blocked by the specific anti-TGF-Pi antibody utilized. All -studies were performed with the same anti-TGF-Pi antibody which is isoform-specific. The previously documented inhibitory activity of Dacron on endothelialization may be explained in part by this Dacron-induced release of TGF-l3 from macrophages. The present study does not differentiate among a Dacron-induced up-regulation of TGF$ mRNA transcription, TGF-l3 protein synthesis, and/or release of stored intracytoplasmic TGF-P protein. Three TGF-P isoforms have been isolated, with TGF-Pi the most prevalent and well characterized. TGF-Pi is a 25-kDa homodimeric peptide secreted in a latent form which requires proteolysis or acidification before receptor bindingj. Three TGF-I3 receptors have also been identified6. It is not possible from the present study to quantitate secretion of latent TGF-P inasmuch as this assay specifically identifies active TGF-l3. Activation of TGF-l3 generally results from proteolysis in an acidic environment. The mitogenic activity of TGF-l3 on both endothelial cells and smooth muscle cells is complex. Endothelial cell growth in vitro is inhibited by TGF-/3’, endothelial ase is delayed’ and c-myc cell entry into the S expression is decreasedgh. However, endothelial cells cultured in three-dimensional collagen gels are not TGF-l3 inhibited” and TGF-/3 installation into newborn and chorioallantoic membranesi induces mice’l angiogenesis. Higher concentrations of TGF-P however resulted in an inhibition of angiogenesis in these models13. TGF-I3 probably plays a significant role in capillary angiogenesis. Exposure of bovine aortic or capillary endothelial cells to fibroblast growth factor induces a significant increase in plasminogen activator activity, probably essential for basement membrane degradation required to permit endothelial sprout migration in the early phase of new capillary formation. However, high concentrations of TGF-Pi (5-10 rig/ml) cause an inhibition of TGF-B-induced angiogenesis by bovine microvascular endothelial cells cultured on three-dimensional collagen or fibrin gels i4 . Introduction of chondrocyteconditioned media into cultures of endothelial cell monolayers induces increased plasminogen activator inhibitor type 1 (PAI-1) protein and mRNA production, a response blocked by pre-incubation of the media with a neutralizing anti-TGF-l3 antibody”. Chondrocyte/ endothelial cell co-culture significantly inhibits capillary tube formation, this activity also being blocked by anti-TGF-I3 antibodies l6 . These data suggest that the anti-angiogenic properties of cartilage may be mediated by chondrocyte-derived TGF-6 induction of PAIproduction by endothelial cells”. The data collected in the experiment showing TGF-l3 secretion by macrophages exposed to Dacron in vitro should be further investigated by studies to document similar effects in
172
vivo and to determine the relationship of secreted TGF-(3 to endothelialization of Dacron prostheses in vivo. The effect of TGF-P on fibroblast and smooth muscle cell growth shows a bimodal concentration effect. Low concentrations of TGF-I3 stimulate fibroblast and smooth muscle cell proliferation whereas higher concentrations inhibit it. This is probably due to the ability of low TGF-P concentrations to up-regulate plateletderived growth factor (PDGF) transcription in fibroblasts and smooth muscle cells whereas higher concentrations simultaneousill down-regulate PDGF-I3 receptor subunit expression . TGF-0 is also a potent stimulator of extracellular matrix molecular synthesis. Subcutaneous TGF-l3 injection induces collagen synthesis’* and TGF-I3 has been shown to induce gene expression and protein synthesis of collagen types I and III and fibronectin’9-21. Spontaneous endothelialization of Dacron grafts implanted clinically has not been reported. The pannus ingrowth of endothelial cells across the anastomosis is limited to approximately 1 cm in length and this perianastomotic region is characterized by exuberant smooth muscle cell and fibroblast proliferation and collagen deposition. These histologic responses result in a relatively thrombogenic acellular flow surface over the majority of the graft’s length and the development of peri-anastomotic pseudointimal hyperplasia. All of these phenomena may in part be mediated by TGF-P production which could induce both smooth muscle cell and fibroblast proliferation and collagen synthesis as well as inhibit endothelial cell ingrowth. However, extrapolation of the in vitro data presented here directly to the in vivo scenario is not possible. Additional studies are warranted to investigate TGF-P protein and message levels in vivo as a function of graft type, anatomic location vis-a-vis the peri-anastomotic region, time following implantation and cell of origin. These studies are currently in progress using immunohistochemical and in situ hybridization techniques.
Acknowledgements These studies were supported by a grant from the National Institutes of Health, ROl HL 41272. The authors also thank Dr A. Gown for the gift of rabbit macrophage-specific anti-RAM 11 antibody.
References 1.
2. 3.
Foxall TL, Auger KR, Callow AD et al. Adult human endothelial cell coverage of small-caliber Dacron and polytetrafluoroethylene vascular prostheses in r&o. J Surg Res 1986; 41: 158-72. Greisler HP, Schwartz TH, Ellinger J et al. Dacron inhibition of arterial regenerative activity. ] Vusc Surg 1986; 3: 747-56. Yam LY, Li CY, Crosby WH. Cytochemical identification of monocytes and granulocytes. Am J Clin Pathol 1971; 55: 283-90.
CARDIOVASCULAR SURGERY APRIL 1996 VOL 4 NO 2
Dacron stimulation of TGF$ rele%w h’. I? Greisier et al. Bianco C, Pitowski B. Fc and C3 receptors. In: Adams DO, Edelson PJ, Koren HS, eds. Methods for Studying Mononuclear Phugocytes. New York: Academic Press, 1981: 273-328. 5. Assoian RK, Komoriya A, Meyers CA et al. Transforming growth factor-beta in human platelets. Identification of a major storage site, purification, and characterization. / Biol Chem 1983; 258: 7155-60. 6. Barnard JA, Lyons RM, Moses HL. The ceil biology of transforming growth factor f5. Biochim Biophys Actu 1990; 1032: 79-87. 7. Frater-Schroder M, Risau W, Hailmann R et al. Tumor necrosis factor type (Y, a potent inhibitor of endotheiiai ceil growth in vitro, is angiogenic in viuo. Proc Nat1 Acad Sci USA 1987; 84: 5277-81 RL, Twardzik DR, Schwartz SM. Inhibition of 8. Heimark endotheiial regeneration by type-beta transforming growth factor from platelets. Science 1986; 233: 1078-80. 9. Takehara K, LeRoy EC, Grotendorst GR. TGF-B inhibition of endotheiial ceil proliferation: alteration of EGF binding and EGF-induced growth regulatory (competence) gene expression. Cdl 1987; 49: 415-22. 10. Madri JA, Pratt BM, Tucker AM. Phenotypic modulation of rndotheiial ceils by transforming growth factor-beta depends upon the composition and organization of the extraceiiuiar matrix. I Cell Biol 1988; 106: 1375-84. 11. Roberts AB, Spom MB, Assoian RK et al. Transforming growth factor type B: rapid induction of fibrosis and angiogenesis in viva and stimulation of collagen formation in vitro. Proc Nat1 Acad Sci USA 1986; 83: 4167-71. growth factor beta 12. Yang EY, Moses HL. Transforming l-induced changes in cell migration, proliferation, and angiogenesis in the chicken chorioailantoic membrane. ] Cell Biol1990; 111: 731-41. 13. Moses HL, Yang EY, Pietenpol JA. TGF-P stimulation and inhibition of ceil proliferation: new mechanistic insights. Cell 4.
CAfEliOVASCULAA SURGERY APRIL 1SQ6 VOL 4 NO 2
14. 15.
16.
17.
18.
19. 20.
21.
3 990; 63: 245-7. Pepper MJ, Vassaih JD, Orci I. ef a/. Bipbasic rffect of transforming growth factor-B, on in ~itn~ air~iogenesn,. F:xp Crll Res 1993; 204: 356-63. Pepper MS, Montesano R, Orci L et al. Plasmrnogen activatorinhibitor-l is induced in microvascular endotheiial cells by a chondrocyte-derived transforming growrh tactor-beta. Hiochen? Biophys Res Comwnm 1991; 176: 633-8. Pepper MS, Montesano R, Vassalii JD ri c?i (bondrocytrs inhibit endothelial sprout formation m t~itro: evidence for the involvement of a transforming growth factclr-beta. 1 Ckl! Phs?/tol 1991; 146: 170-9. Leof EB, Proper JA, Goustin AS et al. induction ot c-srs mRNA and activity similar to platelet-derived growth [actor by transforming growth factor beta: a proposed model for mdirect mitogenesis involving autocrine activity. i~rr)~; Yzzt! Arad 5’~ USA 1986; 83: 2453-7. Roberts AB, Sporn MB, Assoian RK er al. Transforming growth factor type B: rapid induction of fibrosis and dngiogenesis in viva and stimulation of collagen formation in VI~YO.Pro<. Nat/ Acad Sci USA 1986; 83: 4167-71. ignotz RA, Endo T, Massague J. Regulation of ftbronectin and type I collagen mRNA levels by transforming growth factor-(-l. / Biol Chem 1987; 262: 6443-6. Varga J, Rosenbloom J, Jimenez SA. Transforming growth factor-p (TGF-R) causes a persistent increase in steady-state amounts of type I and type II1 collagen and iihrtmectm mRNAs in normal dermai fibrobiasts. Biochenrrstn* 1987: 247. 597-604. Rossi P, Karsentry G, Roberts AB et JI. A nuclaar tactor I binding site mediates the transcriptianal activation of a type I collagen promoter by transforming growth tnctor-/S. (‘elf 1988; 52: 405-14.
Paper accepted 11 July 1994
173