Inhibition of Accelerated Graft Arteriosclerosis by Gene Transfer of Soluble Fibroblast Growth Factor Receptor-1 in Rat Aortic Transplants
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Departments of Medicine (W.L., A.L., H.M.L., B.J.B.), Pathology (Y.C., W.B. III, R.H.H., P.M., B.J.B.), and Surgery (J.H.B., A.D.), Johns Hopkins University School of Medicine, Baltimore, Md.; Department of Medicine (B.J.B.), Albert Einstein College of Medicine, Bronx, NY; and Genetic Therapy Inc. (A Novartis Company) (J.M.-N., S.C.S.), Gaithersburg, Md.
ABSTRACT
Objective— Because increased fibroblast growth factor-1 (FGF-1) and FGF receptor (FGFR) expression correlate with the development of accelerated graft arteriosclerosis in transplanted human hearts, this study sought to determine whether local gene transfer of soluble FGFR-1, capable of binding both FGF-1 and FGF-2, could blunt the development of accelerated graft arteriosclerosis in the rat aortic transplant model.
Methods and Results— A construct encoding the FGFR-1 ectodomain, capable of neutralizing FGF-2 action, was expressed in rat aortic allografts, using adenoviral gene transfer at the time of transplantation. Neointima formation was inhibited in aortic allografts transduced with soluble FGFR-1, compared with allografts transduced with Null virus.
Conclusions— FGFs play a causal role in the development of accelerated graft arteriosclerosis in the rat aortic transplant model. Targeted interruption of FGF function could potentially reduce neointima formation in patients with heart and kidney transplants.
We explored whether local gene transfer of soluble FGF receptor 1 can blunt the development of accelerated graft arteriosclerosis in rat aortic transplants. After adenoviral gene transfer, sFGFR-1 protein was expressed in endothelium and adventitia. Neointima formation was inhibited in aortic allografts transduced with sFGFR-1, but not Null virus.
Key Words: transplantation ? transplant vasculopathy ? chronic rejection ? neointima
Introduction
Accelerated graft arteriosclerosis (AGA), a process leading to diffuse, concentric neointimal accumulation in arteries and arterioles of transplanted hearts, is the leading cause of graft loss and death in patients who have undergone cardiac transplantation.1 Transplant vasculopathy resembling AGA also occurs in renal transplantation, in which it accounts, at least in part, for the progressive loss of graft function that is often referred to as chronic rejection.2
Fibroblast growth factors (FGFs) participate in the vascular remodeling response to injury.3 Fibroblast growth factor-1 (FGF-1) and FGF-2 lack a signal peptide, but they can be released from injured endothelial cells,4 potentially through a vesicle exocytosis-associated process.5 The FGFs are heparin-binding growth factors; hence, proteoglycans in the extracellular matrix may serve as a reservoir for released FGFs.6 Endothelial and smooth muscle cell (SMC) FGF-2 mRNA abundance increases in rats after intraarterial balloon injury,7 and activated macrophages, often present in the vessel wall after different forms of injury, can express and present FGF-2 to other cells.8 Neutralizing FGF-2 antibodies inhibit neointima formation after mechanical injury of arteries,9,10 and gene transfer of FGF-1 cDNA encoding a secreted form of FGF-1 into porcine arteries stimulate neointima formation.11
With regard to transplant arteriopathy, FGF-1 and FGF-2 mRNA and protein abundance increase in human hearts after transplantation,12 and marked upregulation of full-length FGFR-1 is observed.13 Furthermore, a strong correlation between high levels of FGF-1 expression and the presence of cardiac allograft vasculopathy has been reported in patients after cardiac transplantation.14 It has therefore been suggested that FGF-1, and possibly FGF-2, plays an important role in the development of cardiac AGA. Others argue, also on the basis of expression studies, that platelet-derived growth factor (PDGF) plays a more important role in the development of AGA.15
In transplanted human16 and rat17 kidneys with chronic allograft rejection, FGF-1 mRNA and protein also are induced and are found both in the tubulointerstitial compartment18 and in blood vessels, particularly in areas of inflammation and intimal accumulation.2
FGFR-1, a transmembrane receptor tyrosine kinase, is expressed as multiple isoforms that are produced through variations in RNA splicing.19,20 FGF-1 (acidic FGF) and FGF-2 (basic FGF) both interact with FGFR-1.19 The FGFs stimulate proliferation and migration of endothelial cells, fibroblasts, and, to a lesser extent, vascular SMCs.21–23 The proliferative actions of FGF-1 and FGF-2 are mediated, at least in part, through activation of the extracellular signal regulated kinase (ERK) 1/2 mitogen-activated protein kinase pathway21 and require a prolonged stimulus, whereas cell migration occurs in response to a brief stimulus with FGFs and involves activation of Src.22
A soluble form of the FGFR-1, lacking transmembrane and cytoplasmic domains, has been found in vivo and has been postulated to regulate the availability of FGF to cell-surface receptors.24 Soluble FGFR-1 ectodomain can bind FGF-1 and FGF-2 with high affinity,25–27 and its overexpression has been used in the exploration of the role of FGFs in embryonic development.28,29 Expression of soluble FGFR-1, after delivery of the cDNA construct into cells by adenoviral gene transfer, profoundly inhibits vascular SMC growth in vitro.30
The current study explored the question whether a causal role for FGFs could be established in the process of AGA in vivo. A soluble FGFR-1 ectodomain cDNA was delivered into rat aortae just before transplantation. Expression of the soluble FGFR-1 (sFGFR-1) protein was observed in the vessels, and there was significant inhibition of AGA as late as 3 months after transplantation.
Methods
Summary
(For detailed methodology, please see http://atvb.ahajournals.org.) In brief, a cDNA construct encoding the extracellular domain of FGFR-1 (sFGFR-1), without transmembrane or cytoplasmic kinase domain, was created and expressed stably in cultured rat aortic endothelial cells (RAECs) and human embryonic kidney 293 (HEK 293) cells. Epitope-tagged sFGFR-1 construct was then incorporated into a "gutless" adenoviral vector (deleted in proteins E1, E2a, and E3) by homologous recombination. This viral construct is referred to as Av-flag-sFGFR-1. The Av-flag-sFGFR-1 was expressed in cultured cells and in rat aorta by adenoviral gene transfer. Supernatants of cultured cells were evaluated for secreted sFGFR-1 by Western blot analysis, by FGF-2 affinity cross-linking, and for their ability to inhibit the biological activity of FGF-2. Gene transfer into aorta was evaluated using Av-LacZ. Expression of sFGFR-1 mRNA and protein in rat aorta was evaluated by RT-PCR and by immunohistochemistry for the epitope tag, respectively.
The effect of sFGFR-1 on AGA was evaluated in the rat aortic transplant model using the inbred rat strains DA and PVG as donors and recipients, respectively. Gene transfer of Av-flag-sFGFR-1, Av-Null, and buffer was performed by infusion of virus into donor aorta and incubation for 20 minutes immediately before transplantation. The allografts were harvested at 5, 30, 60, and 90 days after transplantation. Immunohistochemistry for smooth muscle actin (a-SM actin), rat macrophage (ED-1), and FGF-2 were performed at each time point. Morphometric analysis was used to quantitate neointima accumulation at 30, 60, and 90 days.
Statistical Analysis
To determine a treatment (group) ±time effect on neointima formation, repeated-measures ANOVA was performed. ANOVA included morphometric data for all rat aortas harvested at 30, 60, and 90 days after transplantation.
Results
FGF-2 Expression in Rat Aortic Allografts
Immunohistochemical analysis with the rat macrophage specific ED-1 antibody and with anti–FGF-2 antibody was performed on rat aortic interposition grafts (Figure I, available online at http://atvb.ahajournals.org). In rat aortic isografts (DA to DA transplant) 5 days after transplantation, FGF-2 protein expression was below the level of detection (Figure IA). In the isografts, ED-1-reactive macrophages were present in adventitia but not in the intima (Figure IB). In rat aortic allografts 5 days after transplantation, FGF-2 and ED-1 immunoreactivity were observed both in the adventitia and in the intima (Figure IC and ID). Moderate to severe macrophage infiltration, most abundant in the aortic adventitia, continued in the aortic allografts at 30, 60 (Figure IF), and 90 days after transplantation. FGF-2 expression was consistently observed in the adventitia up to 60 days after transplantation (Figure IE), but expression waned by 90 days after transplantation (data not shown). Significant neointima formation was observed in aortic allografts 60 days after transplantation (Figure I).
In Vitro Expression and Function of sFGFR-1
In RAEC lines stably transfected with sFGFR-1 cDNA, high-level sFGFR-1 mRNA expression was observed by Northern blot analysis (data not shown). An 70 kDa protein complex labeled with FGF-2 was observed after affinity crosslinking in supernatants of RAECs stably transfected with sFGFR-1 cDNA but not in the supernatant of RAECs stably transfected with empty vector (Figure 1A). The supernatant of RAECs stably transfected with the sFGFR-1 cDNA inhibited FGF-2–stimulated 3T3 fibroblast proliferation, an effect not observed with supernatant of RAECs transfected with vector alone (Figure 1B). The 3T3 fibroblasts continued to proliferate in response to 10% FBS and 10 ng/mL PDGF in the presence of supernatant from RAECs stably transfected with the sFGFR-1 cDNA (not shown).
Figure 1. sFGFR-1 binds FGF-2 with high affinity, and inhibits FGF-2 stimulated 3T3 fibroblast proliferation. A, Affinity cross-linking of secreted proteins with FGF-2. Lanes 1 and 2, Supernatant from RAECs stably transfected with sFGFR-1. Lanes 3 and 4, Supernatant from RAECs stably transfected with empty vector. Affinity cross-linking to FGF-2(141 pM) was performed in the presence (lanes 1 and 3) or absence (lanes 2 and 4) of excess (28 nM) unlabeled FGF-2. B, Effect of RAEC supernatants on FGF-2-stimulated fibroblast proliferation. 3T3 cells were treated with increasing concentrations of FGF-2 in the presence of medium alone,1 in the presence of supernatant from RAECs stably transfected with empty vector,2 and in the presence of supernatants from RAECs transfected with sFGFR-1.3 The sFGFR-1 containing supernatant completely abolished 3T3 fibroblast proliferation in response to FGF-2. C, Western blot of HEK 293 lysates (lanes 1 through 6) and supernatant (lane 7) and of RAEC lysates (lanes 10 through 13) and supernatants (lanes 8 and 9) with anti-flag monoclonal antibody. Lanes 1 through 6, HEK 293 lysates from cells transduced with Av-Flag-sFGFR-1 at multiplicities of infection (MOIs) of 0, 50, 100, 200, 500, and 1000, respectively; lane 7, concentrated conditioned medium from HEK 293 cells transduced with Av-Flag-sFGFR-1 at MOI of 500; lanes 8 and 9, concentrated conditioned medium from RAECs transduced with Av-Flag-sFGFR-1 at MOIs of 20 000 and 10 000; lanes 10 through 13, lysates from RAECs transduced with Av-Flag-sFGFR-1 at MOIs of 0, 5, 1000, 10 000, and 20 000, respectively. D, Inhibition of ERK 1/2 phosphorylation by sFGFR-1 containing supernatants. Quiescent 3T3 fibroblasts were exposed for 5 minutes to concentrated supernatants from HEK 293 cells transduced with Av-Null (MOI 500, lanes 1 and 3) or Av-sFGFR-1 (MOI 500, lanes 2 and 4) in the absence (lanes 1 and 2) or presence (lanes 3 and 4) of FGF-2 (5 ng/mL). Cell lysates were probed with anti-phospho ERK 1/2 (top) and anti-ERK 1/2 antibodies. Compared with control, basal ERK 1/2 phosphorylation was greater in cells exposed to sFGFR-1, and FGF-2 stimulated ERK 1/2 phosphorylation was blunted (representative of 3 separate experiments).
The N-terminal endogenous signal peptide of sFGFR-1 was replaced, in frame, with preprotrypsinogen signal peptide/flag epitope tag. The Av adenoviral vector was generated with this construct. Expression of the recombinant flag-sFGFR-1 from the Av vector was observed by Western blot analysis using anti-flag antibody in cell lysates and supernatants of HEK 293 cells and RAECs (Figure 1C). The protein produced from RAECs migrated with a slightly higher molecular mass than sFGFR-1 expressed in HEK 293 cells, presumably because of differences in glycosylation. The flag-sFGFR-1 produced from RAEC-bound FGF-2 (data not shown), similar to findings with the native sFGFR-1 ectodomain, and supernatants from HEK 293 cells (Figure 1D) and RAECs transfected in vitro with Av-flag-sFGFR-1 reduced FGF-2 (0.5 ng/mL)-stimulated ERK 1/2 tyrosine phosphorylation in 3T3 fibroblasts. These in vitro studies show that sFGFR-1 is secreted from cells expressing the sFGFR-1 cDNA, can bind FGF-2, and can inhibit FGF-2–stimulated cell proliferation and signaling.
Expression of LacZ and Flag-sFGFR-1 in Rat Aorta in Vivo
Pilot experiments showed that endothelial and adventitial gene transfer was achieved with 109 plaque-forming units (PFUs) of Av1-LacZ (Figure 2A). For experiments with Av-flag-sFGFR-1, aortas were therefore transduced with 3 to 5x109 PFU/aortic segment. Expression of sFGFR-1 mRNA in rat aortic allografts was evaluated 5 days after transplantation by RT-PCR, using nested primers specific for the flag-sFGFR-1 construct. Figure 2B shows expression of sFGFR-1 mRNA in each of 4 aortic allografts transduced with Av-flag-sFGFR-1. No expression was seen in sham-transduced or Av-Null–transduced aorta. Evaluation of aortic allografts on 5 days after transplantation, using a monoclonal antibody directed against the flag epitope tag, showed sFGFR-1 protein expression in the endothelium (Figure 2CI) and in the adventitia (Figure 2CII) of the aorta; no expression was observed in aortas transduced with Av-Null virus (Figure 2CIV). Flag-sFGFR-1 detected in vivo appears largely nucleus-associated, although brown reaction product is also seen surrounding the cells (Figure 2CII, inset). HEK 293 cells transduced in vitro and shown to secrete sFGFR-1 (Figure 1C) similarly show significant nuclear flag-sFGFR-1 immunoreactivity (Figure 2CIII). Adventitial gene transfer likely reflects virus spillage during excision and storage of the transduced aorta before implantation into the recipient. Expression of flag-sFGFR-1 protein could not be demonstrated 30 days after transplantation.
Figure 2. Expression of sFGFR-1 mRNA in rat aortic transplants A, Efficiency of Av1-LacZ gene transfer into rat aorta: Av1-LacZ was instilled into the aorta at 5x107 PFU (I), 5x108 PFU (II), and 109 PFU (III and IV). Aortas were harvested 48 hours after transduction followed by visualization of LacZ. IV represents a magnification of the area enclosed in the box in III representative of 2 separate experiments. B, Nested RT-PCR: for all samples, except the pCMV1-flag-sFGFR-1 vector template, RNA was purified and reverse transcribed (±reverse transcriptase). The nested 5' PCR primers recognize sequence for the preprotrypsinogen signal peptide and the flag epitope tag, respectively; the 3' primers detect the native sFGFR-1 sequence. A PCR product of 850 bp was amplified from the control vector (pCMV1-flag-sFGFR-1). A product of the same size was amplified from 4 different rat aortic specimens 5 days after transduction with Av-flag sFGFR-1 but not from aortic transplants transduced with Av-Null or from native rat aorta. No product was obtained from RNA template when reverse transcriptase was omitted from the cDNA synthesis step. An appropriate-sized product was also amplified from HEK 293 cells transduced with Av-flag sFGFR-1 but not from HEK 293 cells transduced with Av-LacZ or from untransduced cells. C, Immunohistochemistry was performed using a mouse monoclonal anti-flag antibody. CI and CII, Aortic allografts transduced with Av-sFGFR-1 5 days after transplantation. Endothelial cells and adventitial cells show flag immunoreactivity in the aortas transduced with Av-flag-sFGFR-1. Reaction product was concentrated in cells. CIII, In Av-sFGFR-1-transduced HEK 293 cells that secrete sFGFR-1, flag immunoreactivity was similarly cell-associated. CIV, No immunoreactivity in the aorta transduced with Av-Null.
Effect of Av Flag-sFGFR-1 Infection on Neointima Formation After Aortic Transplantation
Aortas transduced with Av-flag-sFGFR-1, Av-Null, or sham were harvested 30, 60, or 90 to 100 days after transplantation for evaluation of neointima formation (Figure 3). In all transplanted aortas, most medial SMCs and medial -SM actin immunoreactivity were lost by day 30 after transplantation (Figure II, available online at http://atvb.ahajournals.org). Consistent with medial SMC loss, the aortic media area decreased significantly as a function of time after transplantation (P =0.011; ANOVA) (Figure 3B), but the decrement in media area did not differ between treatment groups. A dense band of -SM actin-immunoreactive cells was observed on the abluminal side of the media in all allografts at day 30. Adventitial -SM actin-immunoreactive cells were also observed on days 60 and 90 but to a lesser degree than at day 30 (Figure 3A). A neointima consisting of inflammatory cells ±-SM actin-immunoreactive cells was observed in some allografts as early as day 30 (Figures II and 3A). A sizable neointima consisting predominantly of -SM actin-positive cells was observed at days 60 and 90 in every sham-transduced and Av-Null-transduced allograft (Figures II and 3A). In contrast, in 2 of 6 and 3 of 4 allografts transduced with Av-sFGFR-1 and evaluated on days 60 and 90, respectively, no -SM actin-reactive neointima was observed (Figure 3A, 60F and 90F), and only small areas of intimal inflammatory cells were seen (Figure II, cED1). Calculation of neointima to media area ratios showed that on average, the neointima formation was blunted significantly in allografts transduced with Av-flag sFGFR-1 when compared with Av-Null-transduced allografts (P =0.0133; ANOVA). No differences in neointima area were observed in Av-Null–transduced and sham-transduced aortic allografts. No significant difference in media area was found between the groups (Figure 3B).
Figure 3. Neointima formation in rat aortic tranplants as a function of time. A, Immunohistochemistry was performed using a mouse monoclonal anti-SM actin antibody. Aortas were harvested on day 30 (30), 60 (60), or day 90 (90) after transplantation. Aortas were either sham-transduced with saline (S) or transduced with Av-Null (N) or with Av-flag sFGFR-1 (F). -SM actin-immunoreactive neointima is observed in sham-transduced (60S, 90S) and Av-null-transduced (60N, 90N) but not Av-flag sFGFR-1-transduced (60F, 90F) aortic grafts on days 60 and 90 after transplantation. B, Mean intima:media ratio and media area as a function of time. Top, Intima:Media ratio in aortic grafts of sham-transduced (, n=4, 3, and 8, on day 30, 60, and 90, respectively); Av-Null-transduced ( n=3, 4, and 4, on day 30, 60, and 90, respectively), and sFGFR-1-transduced (? n=5, 6, and 4 on day 30, 60, and 90, respectively) aortic grafts (mean±SEM). B, Media areas in the same groups (mean±SEM).
Discussion
The current study was undertaken to determine whether local inhibition of FGF action suppresses the development of neointima in rat aortic transplants in vivo. Using adenovirus fully deleted in E1, E2a, and E3, effective gene transfer into aortic segments was achieved before transplantation, and expression of soluble FGFR-1 ectodomain was observed for at least the first 5 days after transplantation. Significant inhibition of neointima formation was observed in aortic allografts transduced with Av-flag-FGFR-1 compared with allografts transduced with Av-Null or sham. Taken together with the findings in vitro that the protein produced from the sFGFR-1 construct binds FGF-2 with high affinity and inhibits FGF-2–stimulated proliferation of fibroblasts, we conclude that FGFs play a causal role in the development of neointima in this rat aortic transplant model of accelerated graft arteriosclerosis. These findings, in conjunction with data previously published by others showing FGF-1 expression in blood vessels undergoing accelerated graft arteriosclerosis,12,14,16–18 raise the possibility that effective interruption of FGF function could also be useful in reducing neointima formation in transplanted human organs.
The model used here was one of rat aortic transplantation, with DA and PVG strains serving as donors and recipients, respectively. With limited immunosuppression consisting of cyclosporine treatment for the first 5 days after transplantation, a neointima developed before 60 days in every aortic allograft not transduced with Av-flag-sFGFR-1. In DA to DA isografts, neointima formation was not observed (data not shown). Hence, immune-mediated injury was necessary for the development of neointima in this model, and the neointima developed predictably within 60 days after transplantation.
Between 5 and 30 days after transplantation, there was a loss of medial SMCs in all aortic transplants, including those with Av-flag-sFGFR-1 gene transfer, with the development of an adventitial ring of -SM actin-immunoreactive cells. CD8+ T lymphocyte-mediated apoptosis, and consequent loss of medial SMCs, has previously been observed in a similar rat aortic allograft model of accelerated graft arteriosclerosis.31 In that study, preservation of medial SMCs in rats depleted of CD8 T lymphocytes with the Ox8 monoclonal antibody did not reduce neointima formation. In our study, occasional areas of medial SMC preservation were found (Figure 3A, 60F; Figure II, cVSM). These were no more frequent in transplants transduced with Av-flag-sFGFR-1 compared with those transduced with Av-Null or sham with saline. Consistent with that observation, the reduction in media size as a function of time after transplantation, presumably because of medial SMC loss, did not differ between the groups (Figure 3B). Aortas in which medial SMCs were absent were still protected from the development of neointima when transduced with Av-flag-sFGFR-1 (Figure 3A, 90F). Therefore, reduced neointima in rat aortic allografts transduced with Av-flag-sFGFR-1 is not explained by preservation of medial SMCs. We observed accumulation of adventitial and intimal macrophages in the aortic grafts (Figure I) as early as 5 days after transplantation. Significant macrophage infiltration in the adventitia persisted through 90 days after transplantation. That T-lymphocyte and macrophage infiltration play a role in the immune response to cardiac transplantation in humans and in carotid artery allografts rodents is well accepted.32–34 FGF-2 expression was observed in the intima and adventitia of aortic allografts as early as day 5 and persisted in the adventitia at day 60 after transplantation (Figure I). FGF-2 expression was not observed in DA to DA isografts, nor did a neointima form in isografts.
We chose the soluble, kinase-deficient FGFR-1 to interrupt FGF action for a several reasons. This receptor is capable of binding FGF-1 and FGF-2 with high affinity26 and was previously reported to inhibit FGF actions in vitro and in vivo.26,28,30,35 We found that the sFGFR-1 protein was expressed and secreted in vitro (Figure 1A and 1C), competitively inhibited FGF-2 binding to fibroblasts (data not shown), and inhibited FGF-2 function in fibroblasts in vitro (Figure 1B and 1D).
Local gene transfer of sFGFR-1 was done immediately before harvesting of the donor aorta. Expression of sFGFR-1 mRNA and protein, the latter in the endothelium and adventitia of the transplanted aorta, was observed on day 5 after transplantation (Figure 2). The sFGFR-1 protein was not observed in aortas harvested at day 30. Hence, expression was either limited to the early period after transplant or below the level of detection at the later time points. Whether endothelial or adventitial sFGFR-1 resulted in inhibition of neointima accumulation cannot be resolved by this study.
Neointima formation was significantly reduced in vessels pretreated with Av-flag-sFGFR-1 gene transfer compared with that observed in sham-transduced aortas and in aortas transduced with Av-Null, with clear separation of the sFGFR-1 group from the two control groups by day 90 (Figure 3B). It might be argued that the dissociation in time of sFGFR-1 expression early from the apparent late inhibitory effect on neointima accumulation rules out a therapeutic effect of sFGFR-1, particularly because FGF-2 expression continues to be observed. However, a strong early inhibitory effect on neointimal cell accumulation could well produce large differences in the size of the neointima late, even if the inhibitory effect of sFGFR-1 waned. Exponential increments in cell number after only a short period of differential cell growth in one group would strongly magnify early differences with time. Because expression of sFGFR-1 was observed only during the period immediately after transplant, we speculate that an initial step in the process of neointima formation was inhibited by sFGFR-1, possibly the migration of adventitial myofibroblasts into the intimal location. It is also possible that interruption of FGF function was longer lived than detection of the flag-sFGR-1 protein and that the mechanism is related to inhibition of intimal SMC with or without myofibroblast proliferation. Although we have not identified the precise mechanism of action, the data show that sFGFR-1, expressed locally in transplanted rat aorta, can significantly inhibit neointima formation in this model of AGA.
Taken together with previous work showing expression of FGFs in vessels of transplanted hearts and kidneys,12,16,17 and a correlation between the level of FGF-1 expression and the development of AGA in human heart transplants,14 our work supports the view that FGFs participate causally in the development of AGA. If these findings also hold in human studies, inhibition of FGF action either by use of soluble receptor ectodomain or by inhibition with specific receptor antagonists could potentially serve to prevent or ameliorate AGA in human solid organ transplantation.
Acknowledgments
This work was supported by a grant from Novartis Pharma and by National Institutes of Health (NIH) grant RO1 DK50670 (to B.J.B.), NIH grant HL09867 (to J.B.), and NIH grant F32HL09263 (to A.D.).
References
Weis M, von Scheidt W. Coronary artery disease in the transplanted heart. Annu Rev Med. 2000; 51: 81–100.
Paul LC. Pathophysiology of chronic renal allograft rejection. Transplant Proc. 1999; 31: 2715–2716.
Reidy MA, Lindner V. Basic FGF and growth of arterial cells. Ann N Y Acad Sci. 1991; 638: 290–299.
Gajdusek CM, Carbon S. Injury-induced release of basic fibroblast growth factor from bovine aortic endothelium. J Cell Physiol. 1989; 139: 570–579.
Carreira CM, LaVallee TM, Tarantini F, Jackson A, Lathrop JT, Hampton B, Burgess WH, Maciag T. S100A13 is involved in the regulation of fibroblast growth factor-1 and p40 synaptotagmin-1 release in vitro. J Biol Chem. 1998; 273: 22224–22231.
Moscatelli D. Basic fibroblast growth factor (bFGF) dissociates rapidly from heparan sulfates but slowly from receptors. Implications for mechanisms of bFGF release from pericellular matrix. J Biol Chem. 1992; 267: 25803–25809.
Lindner V, Reidy MA. Expression of basic fibroblast growth factor and its receptor by smooth muscle cells and endothelium in injured rat arteries. An en face study. Circ Res. 1993; 73: 589–595.
Clasper S, Vekemans S, Fiore M, Plebanski M, Wordsworth P, David G, Jackson DG. Inducible expression of the cell surface heparan sulfate proteoglycan syndecan-2 (fibroglycan) on human activated macrophages can regulate fibroblast growth factor action. J Biol Chem. 1999; 274: 24113–24123.
Lindner V, Reidy MA. Proliferation of smooth muscle cells after vascular injury is inhibited by an antibody against basic fibroblast growth factor. Proc Natl Acad Sci U S A. 1991; 88: 3739–3743.
Nguyen HC, Steinberg BM, LeBoutillier M III, Baumann FG, Rifkin DB, Grossi EA, Galloway AC. Suppression of neointimal lesions after vascular injury: a role for polyclonal anti-basic fibroblast growth factor antibody. Surgery. 1994; 116: 456–461;discussion 461–462.
Nabel EG, Yang ZY, Plautz G, Forough R, Zhan X, Haudenschild CC, Maciag T, Nabel GJ. Recombinant fibroblast growth factor-1 promotes intimal hyperplasia and angiogenesis in arteries in vivo. Nature. 1993; 362: 844–846.
Zhao XM, Yeoh TK, Frist WH, Porterfield DL, Miller GG. Induction of acidic fibroblast growth factor and full-length platelet-derived growth factor expression in human cardiac allografts. Analysis by PCR, in situ hybridization, and immunohistochemistry. Circulation. 1994; 90: 677–685.
Zhao XM, Frist WH, Yeoh TK, Miller GG. Modification of alternative messenger RNA splicing of fibroblast growth factor receptors in human cardiac allografts during rejection. J Clin Invest. 1994; 94: 992–1003.
Miller GG, Davis SF, Atkinson JB, Chomsky DB, Pedroso P, Reddy VS, Drinkwater DC, Zhao XM, Pierson RN. Longitudinal analysis of fibroblast growth factor expression after transplantation and association with severity of cardiac allograft vasculopathy. Circulation. 1999; 100: 2396–2399.
Shaddy RE, Hammond EH, Yowell RL. Immunohistochemical analysis of platelet-derived growth factor and basic fibroblast growth factor in cardiac biopsy and autopsy specimens of heart transplant patients. Am J Cardiol. 1996; 77: 1210–1215.
Kerby JD, Verran DJ, Luo KL, Ding Q, Tagouri Y, Herrera GA, Diethelm AG, Thompson JA. Immunolocalization of FGF-1 and receptors in human renal allograft vasculopathy associated with chronic rejection. Transplantation. 1996; 62: 467–475.
Paul LC, Saito K, Davidoff A, Benediktsson H. Growth factor transcripts in rat renal transplants. Am J Kidney Dis. 1996; 28: 441–450.
Kerby JD, Verran DJ, Luo KL, Ding Q, Tagouri Y, Herrera GA, Diethelm AG, Thompson JA. Immunolocalization of FGF-1 and receptors in glomerular lesions associated with chronic human renal allograft rejection. Transplantation. 1996; 62: 190–200.
Plotnikov AN, Hubbard SR, Schlessinger J, Mohammadi M. Crystal structures of two FGF-FGFR complexes reveal the determinants of ligand-receptor specificity. Cell. 2000; 101: 413–424.
Johnson DE, Williams LT. Structural and functional diversity in the FGF receptor multigene family. Adv Cancer Res. 1993; 60: 1–41.
Landgren E, Klint P, Yokote K, Claesson-Welsh L. Fibroblast growth factor receptor-1 mediates chemotaxis independently of direct SH2-domain protein binding. Oncogene. 1998; 17: 283–291.
LaVallee TM, Prudovsky IA, McMahon GA, Hu X, Maciag T. Activation of the MAP kinase pathway by FGF-1 correlates with cell proliferation induction while activation of the Src pathway correlates with migration. J Cell Biol. 1998; 141: 1647–1658.
Olson NE, Kozlowski J, Reidy MA. Proliferation of intimal smooth muscle cells. Attenuation of basic fibroblast growth factor 2-stimulated proliferation is associated with increased expression of cell cycle inhibitors. J Biol Chem. 2000; 275: 11270–11277.
Hanneken A, Ying W, Ling N, Baird A. Identification of soluble forms of the fibroblast growth factor receptor in blood. Proc Natl Acad Sci U S A. 1994; 91: 9170–9174.
Hanneken A, Maher PA, Baird A. High affinity immunoreactive FGF receptors in the extracellular matrix of vascular endothelial cells: implications for the modulation of FGF-2. J Cell Biol. 1995; 128: 1221–1228.
Ueno H, Gunn M, Dell K, Tseng A Jr, Williams L. A truncated form of fibroblast growth factor receptor 1 inhibits signal transduction by multiple types of fibroblast growth factor receptor. J Biol Chem. 1992; 267: 1470–1476.
Chellaiah A, Yuan W, Chellaiah M, Ornitz DM. Mapping ligand binding domains in chimeric fibroblast growth factor receptor molecules. Multiple regions determine ligand binding specificity. J Biol Chem. 1999; 274: 34785–34794.
Amaya E, Musci TJ, Kirschner MW. Expression of a dominant negative mutant of the FGF receptor disrupts mesoderm formation in Xenopus embryos. Cell. 1991; 66: 257–270.
McFarlane S, Zuber ME, Holt CE. A role for the fibroblast growth factor receptor in cell fate decisions in the developing vertebrate retina. Development. 1998; 125: 3967–3975.
Yukawa H, Miyatake SI, Saiki M, Takahashi JC, Mima T, Ueno H, Nagata I, Kikuchi H, Hashimoto N. In vitro growth suppression of vascular smooth muscle cells using adenovirus-mediated gene transfer of a truncated form of fibroblast growth factor receptor. Atherosclerosis. 1998; 141: 125–132.
Legare JF, Issekutz T, Lee TD, Hirsch G. CD8+ T lymphocytes mediate destruction of the vascular media in a model of chronic rejection. Am J Pathol. 2000; 157: 859–865.
Hruban RH, Kasper EK, Gaudin PB, Baughman KL, Baumgartner WA, Reitz BA, Hutchins GM. Severe lymphocytic endothelialitis associated with coronary artery spasm in a heart transplant recipient. J Heart Lung Transplant. 1992; 11: 42–47.
Moons L, Shi C, Ploplis V, Plow E, Haber E, Collen D, Carmeliet P. Reduced transplant arteriosclerosis in plasminogen-deficient mice. J Clin Invest. 1998; 102: 1788–1797.
Sata M, Luo Z, Walsh K. Fas ligand overexpression on allograft endothelium inhibits inflammatory cell infiltration and transplant-associated intimal hyperplasia. J Immunol. 2001; 166: 6964–6971.
Saiki M, Mima T, Takahashi JC, Tani S, Yukawa H, Ueno H, Mikawa T, Itoh N, Kikuchi H, Hashimoto N, Miyatake S. Adenovirus-mediated gene transfer of a truncated form of fibroblast growth factor receptor inhibits growth of glioma cells both in vitro and in vivo. J Neurooncol. 1999; 44: 195–203.(Wensheng Luo; Ailian Liu;)
ABSTRACT
Objective— Because increased fibroblast growth factor-1 (FGF-1) and FGF receptor (FGFR) expression correlate with the development of accelerated graft arteriosclerosis in transplanted human hearts, this study sought to determine whether local gene transfer of soluble FGFR-1, capable of binding both FGF-1 and FGF-2, could blunt the development of accelerated graft arteriosclerosis in the rat aortic transplant model.
Methods and Results— A construct encoding the FGFR-1 ectodomain, capable of neutralizing FGF-2 action, was expressed in rat aortic allografts, using adenoviral gene transfer at the time of transplantation. Neointima formation was inhibited in aortic allografts transduced with soluble FGFR-1, compared with allografts transduced with Null virus.
Conclusions— FGFs play a causal role in the development of accelerated graft arteriosclerosis in the rat aortic transplant model. Targeted interruption of FGF function could potentially reduce neointima formation in patients with heart and kidney transplants.
We explored whether local gene transfer of soluble FGF receptor 1 can blunt the development of accelerated graft arteriosclerosis in rat aortic transplants. After adenoviral gene transfer, sFGFR-1 protein was expressed in endothelium and adventitia. Neointima formation was inhibited in aortic allografts transduced with sFGFR-1, but not Null virus.
Key Words: transplantation ? transplant vasculopathy ? chronic rejection ? neointima
Introduction
Accelerated graft arteriosclerosis (AGA), a process leading to diffuse, concentric neointimal accumulation in arteries and arterioles of transplanted hearts, is the leading cause of graft loss and death in patients who have undergone cardiac transplantation.1 Transplant vasculopathy resembling AGA also occurs in renal transplantation, in which it accounts, at least in part, for the progressive loss of graft function that is often referred to as chronic rejection.2
Fibroblast growth factors (FGFs) participate in the vascular remodeling response to injury.3 Fibroblast growth factor-1 (FGF-1) and FGF-2 lack a signal peptide, but they can be released from injured endothelial cells,4 potentially through a vesicle exocytosis-associated process.5 The FGFs are heparin-binding growth factors; hence, proteoglycans in the extracellular matrix may serve as a reservoir for released FGFs.6 Endothelial and smooth muscle cell (SMC) FGF-2 mRNA abundance increases in rats after intraarterial balloon injury,7 and activated macrophages, often present in the vessel wall after different forms of injury, can express and present FGF-2 to other cells.8 Neutralizing FGF-2 antibodies inhibit neointima formation after mechanical injury of arteries,9,10 and gene transfer of FGF-1 cDNA encoding a secreted form of FGF-1 into porcine arteries stimulate neointima formation.11
With regard to transplant arteriopathy, FGF-1 and FGF-2 mRNA and protein abundance increase in human hearts after transplantation,12 and marked upregulation of full-length FGFR-1 is observed.13 Furthermore, a strong correlation between high levels of FGF-1 expression and the presence of cardiac allograft vasculopathy has been reported in patients after cardiac transplantation.14 It has therefore been suggested that FGF-1, and possibly FGF-2, plays an important role in the development of cardiac AGA. Others argue, also on the basis of expression studies, that platelet-derived growth factor (PDGF) plays a more important role in the development of AGA.15
In transplanted human16 and rat17 kidneys with chronic allograft rejection, FGF-1 mRNA and protein also are induced and are found both in the tubulointerstitial compartment18 and in blood vessels, particularly in areas of inflammation and intimal accumulation.2
FGFR-1, a transmembrane receptor tyrosine kinase, is expressed as multiple isoforms that are produced through variations in RNA splicing.19,20 FGF-1 (acidic FGF) and FGF-2 (basic FGF) both interact with FGFR-1.19 The FGFs stimulate proliferation and migration of endothelial cells, fibroblasts, and, to a lesser extent, vascular SMCs.21–23 The proliferative actions of FGF-1 and FGF-2 are mediated, at least in part, through activation of the extracellular signal regulated kinase (ERK) 1/2 mitogen-activated protein kinase pathway21 and require a prolonged stimulus, whereas cell migration occurs in response to a brief stimulus with FGFs and involves activation of Src.22
A soluble form of the FGFR-1, lacking transmembrane and cytoplasmic domains, has been found in vivo and has been postulated to regulate the availability of FGF to cell-surface receptors.24 Soluble FGFR-1 ectodomain can bind FGF-1 and FGF-2 with high affinity,25–27 and its overexpression has been used in the exploration of the role of FGFs in embryonic development.28,29 Expression of soluble FGFR-1, after delivery of the cDNA construct into cells by adenoviral gene transfer, profoundly inhibits vascular SMC growth in vitro.30
The current study explored the question whether a causal role for FGFs could be established in the process of AGA in vivo. A soluble FGFR-1 ectodomain cDNA was delivered into rat aortae just before transplantation. Expression of the soluble FGFR-1 (sFGFR-1) protein was observed in the vessels, and there was significant inhibition of AGA as late as 3 months after transplantation.
Methods
Summary
(For detailed methodology, please see http://atvb.ahajournals.org.) In brief, a cDNA construct encoding the extracellular domain of FGFR-1 (sFGFR-1), without transmembrane or cytoplasmic kinase domain, was created and expressed stably in cultured rat aortic endothelial cells (RAECs) and human embryonic kidney 293 (HEK 293) cells. Epitope-tagged sFGFR-1 construct was then incorporated into a "gutless" adenoviral vector (deleted in proteins E1, E2a, and E3) by homologous recombination. This viral construct is referred to as Av-flag-sFGFR-1. The Av-flag-sFGFR-1 was expressed in cultured cells and in rat aorta by adenoviral gene transfer. Supernatants of cultured cells were evaluated for secreted sFGFR-1 by Western blot analysis, by FGF-2 affinity cross-linking, and for their ability to inhibit the biological activity of FGF-2. Gene transfer into aorta was evaluated using Av-LacZ. Expression of sFGFR-1 mRNA and protein in rat aorta was evaluated by RT-PCR and by immunohistochemistry for the epitope tag, respectively.
The effect of sFGFR-1 on AGA was evaluated in the rat aortic transplant model using the inbred rat strains DA and PVG as donors and recipients, respectively. Gene transfer of Av-flag-sFGFR-1, Av-Null, and buffer was performed by infusion of virus into donor aorta and incubation for 20 minutes immediately before transplantation. The allografts were harvested at 5, 30, 60, and 90 days after transplantation. Immunohistochemistry for smooth muscle actin (a-SM actin), rat macrophage (ED-1), and FGF-2 were performed at each time point. Morphometric analysis was used to quantitate neointima accumulation at 30, 60, and 90 days.
Statistical Analysis
To determine a treatment (group) ±time effect on neointima formation, repeated-measures ANOVA was performed. ANOVA included morphometric data for all rat aortas harvested at 30, 60, and 90 days after transplantation.
Results
FGF-2 Expression in Rat Aortic Allografts
Immunohistochemical analysis with the rat macrophage specific ED-1 antibody and with anti–FGF-2 antibody was performed on rat aortic interposition grafts (Figure I, available online at http://atvb.ahajournals.org). In rat aortic isografts (DA to DA transplant) 5 days after transplantation, FGF-2 protein expression was below the level of detection (Figure IA). In the isografts, ED-1-reactive macrophages were present in adventitia but not in the intima (Figure IB). In rat aortic allografts 5 days after transplantation, FGF-2 and ED-1 immunoreactivity were observed both in the adventitia and in the intima (Figure IC and ID). Moderate to severe macrophage infiltration, most abundant in the aortic adventitia, continued in the aortic allografts at 30, 60 (Figure IF), and 90 days after transplantation. FGF-2 expression was consistently observed in the adventitia up to 60 days after transplantation (Figure IE), but expression waned by 90 days after transplantation (data not shown). Significant neointima formation was observed in aortic allografts 60 days after transplantation (Figure I).
In Vitro Expression and Function of sFGFR-1
In RAEC lines stably transfected with sFGFR-1 cDNA, high-level sFGFR-1 mRNA expression was observed by Northern blot analysis (data not shown). An 70 kDa protein complex labeled with FGF-2 was observed after affinity crosslinking in supernatants of RAECs stably transfected with sFGFR-1 cDNA but not in the supernatant of RAECs stably transfected with empty vector (Figure 1A). The supernatant of RAECs stably transfected with the sFGFR-1 cDNA inhibited FGF-2–stimulated 3T3 fibroblast proliferation, an effect not observed with supernatant of RAECs transfected with vector alone (Figure 1B). The 3T3 fibroblasts continued to proliferate in response to 10% FBS and 10 ng/mL PDGF in the presence of supernatant from RAECs stably transfected with the sFGFR-1 cDNA (not shown).
Figure 1. sFGFR-1 binds FGF-2 with high affinity, and inhibits FGF-2 stimulated 3T3 fibroblast proliferation. A, Affinity cross-linking of secreted proteins with FGF-2. Lanes 1 and 2, Supernatant from RAECs stably transfected with sFGFR-1. Lanes 3 and 4, Supernatant from RAECs stably transfected with empty vector. Affinity cross-linking to FGF-2(141 pM) was performed in the presence (lanes 1 and 3) or absence (lanes 2 and 4) of excess (28 nM) unlabeled FGF-2. B, Effect of RAEC supernatants on FGF-2-stimulated fibroblast proliferation. 3T3 cells were treated with increasing concentrations of FGF-2 in the presence of medium alone,1 in the presence of supernatant from RAECs stably transfected with empty vector,2 and in the presence of supernatants from RAECs transfected with sFGFR-1.3 The sFGFR-1 containing supernatant completely abolished 3T3 fibroblast proliferation in response to FGF-2. C, Western blot of HEK 293 lysates (lanes 1 through 6) and supernatant (lane 7) and of RAEC lysates (lanes 10 through 13) and supernatants (lanes 8 and 9) with anti-flag monoclonal antibody. Lanes 1 through 6, HEK 293 lysates from cells transduced with Av-Flag-sFGFR-1 at multiplicities of infection (MOIs) of 0, 50, 100, 200, 500, and 1000, respectively; lane 7, concentrated conditioned medium from HEK 293 cells transduced with Av-Flag-sFGFR-1 at MOI of 500; lanes 8 and 9, concentrated conditioned medium from RAECs transduced with Av-Flag-sFGFR-1 at MOIs of 20 000 and 10 000; lanes 10 through 13, lysates from RAECs transduced with Av-Flag-sFGFR-1 at MOIs of 0, 5, 1000, 10 000, and 20 000, respectively. D, Inhibition of ERK 1/2 phosphorylation by sFGFR-1 containing supernatants. Quiescent 3T3 fibroblasts were exposed for 5 minutes to concentrated supernatants from HEK 293 cells transduced with Av-Null (MOI 500, lanes 1 and 3) or Av-sFGFR-1 (MOI 500, lanes 2 and 4) in the absence (lanes 1 and 2) or presence (lanes 3 and 4) of FGF-2 (5 ng/mL). Cell lysates were probed with anti-phospho ERK 1/2 (top) and anti-ERK 1/2 antibodies. Compared with control, basal ERK 1/2 phosphorylation was greater in cells exposed to sFGFR-1, and FGF-2 stimulated ERK 1/2 phosphorylation was blunted (representative of 3 separate experiments).
The N-terminal endogenous signal peptide of sFGFR-1 was replaced, in frame, with preprotrypsinogen signal peptide/flag epitope tag. The Av adenoviral vector was generated with this construct. Expression of the recombinant flag-sFGFR-1 from the Av vector was observed by Western blot analysis using anti-flag antibody in cell lysates and supernatants of HEK 293 cells and RAECs (Figure 1C). The protein produced from RAECs migrated with a slightly higher molecular mass than sFGFR-1 expressed in HEK 293 cells, presumably because of differences in glycosylation. The flag-sFGFR-1 produced from RAEC-bound FGF-2 (data not shown), similar to findings with the native sFGFR-1 ectodomain, and supernatants from HEK 293 cells (Figure 1D) and RAECs transfected in vitro with Av-flag-sFGFR-1 reduced FGF-2 (0.5 ng/mL)-stimulated ERK 1/2 tyrosine phosphorylation in 3T3 fibroblasts. These in vitro studies show that sFGFR-1 is secreted from cells expressing the sFGFR-1 cDNA, can bind FGF-2, and can inhibit FGF-2–stimulated cell proliferation and signaling.
Expression of LacZ and Flag-sFGFR-1 in Rat Aorta in Vivo
Pilot experiments showed that endothelial and adventitial gene transfer was achieved with 109 plaque-forming units (PFUs) of Av1-LacZ (Figure 2A). For experiments with Av-flag-sFGFR-1, aortas were therefore transduced with 3 to 5x109 PFU/aortic segment. Expression of sFGFR-1 mRNA in rat aortic allografts was evaluated 5 days after transplantation by RT-PCR, using nested primers specific for the flag-sFGFR-1 construct. Figure 2B shows expression of sFGFR-1 mRNA in each of 4 aortic allografts transduced with Av-flag-sFGFR-1. No expression was seen in sham-transduced or Av-Null–transduced aorta. Evaluation of aortic allografts on 5 days after transplantation, using a monoclonal antibody directed against the flag epitope tag, showed sFGFR-1 protein expression in the endothelium (Figure 2CI) and in the adventitia (Figure 2CII) of the aorta; no expression was observed in aortas transduced with Av-Null virus (Figure 2CIV). Flag-sFGFR-1 detected in vivo appears largely nucleus-associated, although brown reaction product is also seen surrounding the cells (Figure 2CII, inset). HEK 293 cells transduced in vitro and shown to secrete sFGFR-1 (Figure 1C) similarly show significant nuclear flag-sFGFR-1 immunoreactivity (Figure 2CIII). Adventitial gene transfer likely reflects virus spillage during excision and storage of the transduced aorta before implantation into the recipient. Expression of flag-sFGFR-1 protein could not be demonstrated 30 days after transplantation.
Figure 2. Expression of sFGFR-1 mRNA in rat aortic transplants A, Efficiency of Av1-LacZ gene transfer into rat aorta: Av1-LacZ was instilled into the aorta at 5x107 PFU (I), 5x108 PFU (II), and 109 PFU (III and IV). Aortas were harvested 48 hours after transduction followed by visualization of LacZ. IV represents a magnification of the area enclosed in the box in III representative of 2 separate experiments. B, Nested RT-PCR: for all samples, except the pCMV1-flag-sFGFR-1 vector template, RNA was purified and reverse transcribed (±reverse transcriptase). The nested 5' PCR primers recognize sequence for the preprotrypsinogen signal peptide and the flag epitope tag, respectively; the 3' primers detect the native sFGFR-1 sequence. A PCR product of 850 bp was amplified from the control vector (pCMV1-flag-sFGFR-1). A product of the same size was amplified from 4 different rat aortic specimens 5 days after transduction with Av-flag sFGFR-1 but not from aortic transplants transduced with Av-Null or from native rat aorta. No product was obtained from RNA template when reverse transcriptase was omitted from the cDNA synthesis step. An appropriate-sized product was also amplified from HEK 293 cells transduced with Av-flag sFGFR-1 but not from HEK 293 cells transduced with Av-LacZ or from untransduced cells. C, Immunohistochemistry was performed using a mouse monoclonal anti-flag antibody. CI and CII, Aortic allografts transduced with Av-sFGFR-1 5 days after transplantation. Endothelial cells and adventitial cells show flag immunoreactivity in the aortas transduced with Av-flag-sFGFR-1. Reaction product was concentrated in cells. CIII, In Av-sFGFR-1-transduced HEK 293 cells that secrete sFGFR-1, flag immunoreactivity was similarly cell-associated. CIV, No immunoreactivity in the aorta transduced with Av-Null.
Effect of Av Flag-sFGFR-1 Infection on Neointima Formation After Aortic Transplantation
Aortas transduced with Av-flag-sFGFR-1, Av-Null, or sham were harvested 30, 60, or 90 to 100 days after transplantation for evaluation of neointima formation (Figure 3). In all transplanted aortas, most medial SMCs and medial -SM actin immunoreactivity were lost by day 30 after transplantation (Figure II, available online at http://atvb.ahajournals.org). Consistent with medial SMC loss, the aortic media area decreased significantly as a function of time after transplantation (P =0.011; ANOVA) (Figure 3B), but the decrement in media area did not differ between treatment groups. A dense band of -SM actin-immunoreactive cells was observed on the abluminal side of the media in all allografts at day 30. Adventitial -SM actin-immunoreactive cells were also observed on days 60 and 90 but to a lesser degree than at day 30 (Figure 3A). A neointima consisting of inflammatory cells ±-SM actin-immunoreactive cells was observed in some allografts as early as day 30 (Figures II and 3A). A sizable neointima consisting predominantly of -SM actin-positive cells was observed at days 60 and 90 in every sham-transduced and Av-Null-transduced allograft (Figures II and 3A). In contrast, in 2 of 6 and 3 of 4 allografts transduced with Av-sFGFR-1 and evaluated on days 60 and 90, respectively, no -SM actin-reactive neointima was observed (Figure 3A, 60F and 90F), and only small areas of intimal inflammatory cells were seen (Figure II, cED1). Calculation of neointima to media area ratios showed that on average, the neointima formation was blunted significantly in allografts transduced with Av-flag sFGFR-1 when compared with Av-Null-transduced allografts (P =0.0133; ANOVA). No differences in neointima area were observed in Av-Null–transduced and sham-transduced aortic allografts. No significant difference in media area was found between the groups (Figure 3B).
Figure 3. Neointima formation in rat aortic tranplants as a function of time. A, Immunohistochemistry was performed using a mouse monoclonal anti-SM actin antibody. Aortas were harvested on day 30 (30), 60 (60), or day 90 (90) after transplantation. Aortas were either sham-transduced with saline (S) or transduced with Av-Null (N) or with Av-flag sFGFR-1 (F). -SM actin-immunoreactive neointima is observed in sham-transduced (60S, 90S) and Av-null-transduced (60N, 90N) but not Av-flag sFGFR-1-transduced (60F, 90F) aortic grafts on days 60 and 90 after transplantation. B, Mean intima:media ratio and media area as a function of time. Top, Intima:Media ratio in aortic grafts of sham-transduced (, n=4, 3, and 8, on day 30, 60, and 90, respectively); Av-Null-transduced ( n=3, 4, and 4, on day 30, 60, and 90, respectively), and sFGFR-1-transduced (? n=5, 6, and 4 on day 30, 60, and 90, respectively) aortic grafts (mean±SEM). B, Media areas in the same groups (mean±SEM).
Discussion
The current study was undertaken to determine whether local inhibition of FGF action suppresses the development of neointima in rat aortic transplants in vivo. Using adenovirus fully deleted in E1, E2a, and E3, effective gene transfer into aortic segments was achieved before transplantation, and expression of soluble FGFR-1 ectodomain was observed for at least the first 5 days after transplantation. Significant inhibition of neointima formation was observed in aortic allografts transduced with Av-flag-FGFR-1 compared with allografts transduced with Av-Null or sham. Taken together with the findings in vitro that the protein produced from the sFGFR-1 construct binds FGF-2 with high affinity and inhibits FGF-2–stimulated proliferation of fibroblasts, we conclude that FGFs play a causal role in the development of neointima in this rat aortic transplant model of accelerated graft arteriosclerosis. These findings, in conjunction with data previously published by others showing FGF-1 expression in blood vessels undergoing accelerated graft arteriosclerosis,12,14,16–18 raise the possibility that effective interruption of FGF function could also be useful in reducing neointima formation in transplanted human organs.
The model used here was one of rat aortic transplantation, with DA and PVG strains serving as donors and recipients, respectively. With limited immunosuppression consisting of cyclosporine treatment for the first 5 days after transplantation, a neointima developed before 60 days in every aortic allograft not transduced with Av-flag-sFGFR-1. In DA to DA isografts, neointima formation was not observed (data not shown). Hence, immune-mediated injury was necessary for the development of neointima in this model, and the neointima developed predictably within 60 days after transplantation.
Between 5 and 30 days after transplantation, there was a loss of medial SMCs in all aortic transplants, including those with Av-flag-sFGFR-1 gene transfer, with the development of an adventitial ring of -SM actin-immunoreactive cells. CD8+ T lymphocyte-mediated apoptosis, and consequent loss of medial SMCs, has previously been observed in a similar rat aortic allograft model of accelerated graft arteriosclerosis.31 In that study, preservation of medial SMCs in rats depleted of CD8 T lymphocytes with the Ox8 monoclonal antibody did not reduce neointima formation. In our study, occasional areas of medial SMC preservation were found (Figure 3A, 60F; Figure II, cVSM). These were no more frequent in transplants transduced with Av-flag-sFGFR-1 compared with those transduced with Av-Null or sham with saline. Consistent with that observation, the reduction in media size as a function of time after transplantation, presumably because of medial SMC loss, did not differ between the groups (Figure 3B). Aortas in which medial SMCs were absent were still protected from the development of neointima when transduced with Av-flag-sFGFR-1 (Figure 3A, 90F). Therefore, reduced neointima in rat aortic allografts transduced with Av-flag-sFGFR-1 is not explained by preservation of medial SMCs. We observed accumulation of adventitial and intimal macrophages in the aortic grafts (Figure I) as early as 5 days after transplantation. Significant macrophage infiltration in the adventitia persisted through 90 days after transplantation. That T-lymphocyte and macrophage infiltration play a role in the immune response to cardiac transplantation in humans and in carotid artery allografts rodents is well accepted.32–34 FGF-2 expression was observed in the intima and adventitia of aortic allografts as early as day 5 and persisted in the adventitia at day 60 after transplantation (Figure I). FGF-2 expression was not observed in DA to DA isografts, nor did a neointima form in isografts.
We chose the soluble, kinase-deficient FGFR-1 to interrupt FGF action for a several reasons. This receptor is capable of binding FGF-1 and FGF-2 with high affinity26 and was previously reported to inhibit FGF actions in vitro and in vivo.26,28,30,35 We found that the sFGFR-1 protein was expressed and secreted in vitro (Figure 1A and 1C), competitively inhibited FGF-2 binding to fibroblasts (data not shown), and inhibited FGF-2 function in fibroblasts in vitro (Figure 1B and 1D).
Local gene transfer of sFGFR-1 was done immediately before harvesting of the donor aorta. Expression of sFGFR-1 mRNA and protein, the latter in the endothelium and adventitia of the transplanted aorta, was observed on day 5 after transplantation (Figure 2). The sFGFR-1 protein was not observed in aortas harvested at day 30. Hence, expression was either limited to the early period after transplant or below the level of detection at the later time points. Whether endothelial or adventitial sFGFR-1 resulted in inhibition of neointima accumulation cannot be resolved by this study.
Neointima formation was significantly reduced in vessels pretreated with Av-flag-sFGFR-1 gene transfer compared with that observed in sham-transduced aortas and in aortas transduced with Av-Null, with clear separation of the sFGFR-1 group from the two control groups by day 90 (Figure 3B). It might be argued that the dissociation in time of sFGFR-1 expression early from the apparent late inhibitory effect on neointima accumulation rules out a therapeutic effect of sFGFR-1, particularly because FGF-2 expression continues to be observed. However, a strong early inhibitory effect on neointimal cell accumulation could well produce large differences in the size of the neointima late, even if the inhibitory effect of sFGFR-1 waned. Exponential increments in cell number after only a short period of differential cell growth in one group would strongly magnify early differences with time. Because expression of sFGFR-1 was observed only during the period immediately after transplant, we speculate that an initial step in the process of neointima formation was inhibited by sFGFR-1, possibly the migration of adventitial myofibroblasts into the intimal location. It is also possible that interruption of FGF function was longer lived than detection of the flag-sFGR-1 protein and that the mechanism is related to inhibition of intimal SMC with or without myofibroblast proliferation. Although we have not identified the precise mechanism of action, the data show that sFGFR-1, expressed locally in transplanted rat aorta, can significantly inhibit neointima formation in this model of AGA.
Taken together with previous work showing expression of FGFs in vessels of transplanted hearts and kidneys,12,16,17 and a correlation between the level of FGF-1 expression and the development of AGA in human heart transplants,14 our work supports the view that FGFs participate causally in the development of AGA. If these findings also hold in human studies, inhibition of FGF action either by use of soluble receptor ectodomain or by inhibition with specific receptor antagonists could potentially serve to prevent or ameliorate AGA in human solid organ transplantation.
Acknowledgments
This work was supported by a grant from Novartis Pharma and by National Institutes of Health (NIH) grant RO1 DK50670 (to B.J.B.), NIH grant HL09867 (to J.B.), and NIH grant F32HL09263 (to A.D.).
References
Weis M, von Scheidt W. Coronary artery disease in the transplanted heart. Annu Rev Med. 2000; 51: 81–100.
Paul LC. Pathophysiology of chronic renal allograft rejection. Transplant Proc. 1999; 31: 2715–2716.
Reidy MA, Lindner V. Basic FGF and growth of arterial cells. Ann N Y Acad Sci. 1991; 638: 290–299.
Gajdusek CM, Carbon S. Injury-induced release of basic fibroblast growth factor from bovine aortic endothelium. J Cell Physiol. 1989; 139: 570–579.
Carreira CM, LaVallee TM, Tarantini F, Jackson A, Lathrop JT, Hampton B, Burgess WH, Maciag T. S100A13 is involved in the regulation of fibroblast growth factor-1 and p40 synaptotagmin-1 release in vitro. J Biol Chem. 1998; 273: 22224–22231.
Moscatelli D. Basic fibroblast growth factor (bFGF) dissociates rapidly from heparan sulfates but slowly from receptors. Implications for mechanisms of bFGF release from pericellular matrix. J Biol Chem. 1992; 267: 25803–25809.
Lindner V, Reidy MA. Expression of basic fibroblast growth factor and its receptor by smooth muscle cells and endothelium in injured rat arteries. An en face study. Circ Res. 1993; 73: 589–595.
Clasper S, Vekemans S, Fiore M, Plebanski M, Wordsworth P, David G, Jackson DG. Inducible expression of the cell surface heparan sulfate proteoglycan syndecan-2 (fibroglycan) on human activated macrophages can regulate fibroblast growth factor action. J Biol Chem. 1999; 274: 24113–24123.
Lindner V, Reidy MA. Proliferation of smooth muscle cells after vascular injury is inhibited by an antibody against basic fibroblast growth factor. Proc Natl Acad Sci U S A. 1991; 88: 3739–3743.
Nguyen HC, Steinberg BM, LeBoutillier M III, Baumann FG, Rifkin DB, Grossi EA, Galloway AC. Suppression of neointimal lesions after vascular injury: a role for polyclonal anti-basic fibroblast growth factor antibody. Surgery. 1994; 116: 456–461;discussion 461–462.
Nabel EG, Yang ZY, Plautz G, Forough R, Zhan X, Haudenschild CC, Maciag T, Nabel GJ. Recombinant fibroblast growth factor-1 promotes intimal hyperplasia and angiogenesis in arteries in vivo. Nature. 1993; 362: 844–846.
Zhao XM, Yeoh TK, Frist WH, Porterfield DL, Miller GG. Induction of acidic fibroblast growth factor and full-length platelet-derived growth factor expression in human cardiac allografts. Analysis by PCR, in situ hybridization, and immunohistochemistry. Circulation. 1994; 90: 677–685.
Zhao XM, Frist WH, Yeoh TK, Miller GG. Modification of alternative messenger RNA splicing of fibroblast growth factor receptors in human cardiac allografts during rejection. J Clin Invest. 1994; 94: 992–1003.
Miller GG, Davis SF, Atkinson JB, Chomsky DB, Pedroso P, Reddy VS, Drinkwater DC, Zhao XM, Pierson RN. Longitudinal analysis of fibroblast growth factor expression after transplantation and association with severity of cardiac allograft vasculopathy. Circulation. 1999; 100: 2396–2399.
Shaddy RE, Hammond EH, Yowell RL. Immunohistochemical analysis of platelet-derived growth factor and basic fibroblast growth factor in cardiac biopsy and autopsy specimens of heart transplant patients. Am J Cardiol. 1996; 77: 1210–1215.
Kerby JD, Verran DJ, Luo KL, Ding Q, Tagouri Y, Herrera GA, Diethelm AG, Thompson JA. Immunolocalization of FGF-1 and receptors in human renal allograft vasculopathy associated with chronic rejection. Transplantation. 1996; 62: 467–475.
Paul LC, Saito K, Davidoff A, Benediktsson H. Growth factor transcripts in rat renal transplants. Am J Kidney Dis. 1996; 28: 441–450.
Kerby JD, Verran DJ, Luo KL, Ding Q, Tagouri Y, Herrera GA, Diethelm AG, Thompson JA. Immunolocalization of FGF-1 and receptors in glomerular lesions associated with chronic human renal allograft rejection. Transplantation. 1996; 62: 190–200.
Plotnikov AN, Hubbard SR, Schlessinger J, Mohammadi M. Crystal structures of two FGF-FGFR complexes reveal the determinants of ligand-receptor specificity. Cell. 2000; 101: 413–424.
Johnson DE, Williams LT. Structural and functional diversity in the FGF receptor multigene family. Adv Cancer Res. 1993; 60: 1–41.
Landgren E, Klint P, Yokote K, Claesson-Welsh L. Fibroblast growth factor receptor-1 mediates chemotaxis independently of direct SH2-domain protein binding. Oncogene. 1998; 17: 283–291.
LaVallee TM, Prudovsky IA, McMahon GA, Hu X, Maciag T. Activation of the MAP kinase pathway by FGF-1 correlates with cell proliferation induction while activation of the Src pathway correlates with migration. J Cell Biol. 1998; 141: 1647–1658.
Olson NE, Kozlowski J, Reidy MA. Proliferation of intimal smooth muscle cells. Attenuation of basic fibroblast growth factor 2-stimulated proliferation is associated with increased expression of cell cycle inhibitors. J Biol Chem. 2000; 275: 11270–11277.
Hanneken A, Ying W, Ling N, Baird A. Identification of soluble forms of the fibroblast growth factor receptor in blood. Proc Natl Acad Sci U S A. 1994; 91: 9170–9174.
Hanneken A, Maher PA, Baird A. High affinity immunoreactive FGF receptors in the extracellular matrix of vascular endothelial cells: implications for the modulation of FGF-2. J Cell Biol. 1995; 128: 1221–1228.
Ueno H, Gunn M, Dell K, Tseng A Jr, Williams L. A truncated form of fibroblast growth factor receptor 1 inhibits signal transduction by multiple types of fibroblast growth factor receptor. J Biol Chem. 1992; 267: 1470–1476.
Chellaiah A, Yuan W, Chellaiah M, Ornitz DM. Mapping ligand binding domains in chimeric fibroblast growth factor receptor molecules. Multiple regions determine ligand binding specificity. J Biol Chem. 1999; 274: 34785–34794.
Amaya E, Musci TJ, Kirschner MW. Expression of a dominant negative mutant of the FGF receptor disrupts mesoderm formation in Xenopus embryos. Cell. 1991; 66: 257–270.
McFarlane S, Zuber ME, Holt CE. A role for the fibroblast growth factor receptor in cell fate decisions in the developing vertebrate retina. Development. 1998; 125: 3967–3975.
Yukawa H, Miyatake SI, Saiki M, Takahashi JC, Mima T, Ueno H, Nagata I, Kikuchi H, Hashimoto N. In vitro growth suppression of vascular smooth muscle cells using adenovirus-mediated gene transfer of a truncated form of fibroblast growth factor receptor. Atherosclerosis. 1998; 141: 125–132.
Legare JF, Issekutz T, Lee TD, Hirsch G. CD8+ T lymphocytes mediate destruction of the vascular media in a model of chronic rejection. Am J Pathol. 2000; 157: 859–865.
Hruban RH, Kasper EK, Gaudin PB, Baughman KL, Baumgartner WA, Reitz BA, Hutchins GM. Severe lymphocytic endothelialitis associated with coronary artery spasm in a heart transplant recipient. J Heart Lung Transplant. 1992; 11: 42–47.
Moons L, Shi C, Ploplis V, Plow E, Haber E, Collen D, Carmeliet P. Reduced transplant arteriosclerosis in plasminogen-deficient mice. J Clin Invest. 1998; 102: 1788–1797.
Sata M, Luo Z, Walsh K. Fas ligand overexpression on allograft endothelium inhibits inflammatory cell infiltration and transplant-associated intimal hyperplasia. J Immunol. 2001; 166: 6964–6971.
Saiki M, Mima T, Takahashi JC, Tani S, Yukawa H, Ueno H, Mikawa T, Itoh N, Kikuchi H, Hashimoto N, Miyatake S. Adenovirus-mediated gene transfer of a truncated form of fibroblast growth factor receptor inhibits growth of glioma cells both in vitro and in vivo. J Neurooncol. 1999; 44: 195–203.(Wensheng Luo; Ailian Liu;)