Differential Healing After Sirolimus, Paclitaxel, and Bare Metal Stent Placement in Combination With Peroxisome Proliferator-Activator Receptor γ Agonists
Requirement for mTOR/Akt2 in PPARγ Activation
Rationale: Sirolimus-eluting coronary stents (SESs) and paclitaxel-eluting coronary stents (PESs) are used to reduce restenosis but have different sites of action. The molecular targets of sirolimus overlap with those of the peroxisome proliferator-activated receptor (PPAR)γ agonist rosiglitazone (RSG) but the consequence of this interaction on endothelialization is unknown.
Objective: Using the New Zealand white rabbit iliac model of stenting, we examined the effects of RSG on SESs, PESs, and bare metal stents endothelialization.
Methods and Results: Animals receiving SESs, PESs, or bare metal stents and either RSG (3 mg/kg per day) or placebo were euthanized at 28 days, and arteries were evaluated by scanning electron microscopy. Fourteen-day organ culture and Western blotting of iliac arteries and tissue culture experiments were conducted. Endothelialization was significantly reduced by RSG in SESs but not in PESs or bare metal stents. Organ culture revealed reduced vascular endothelial growth factor in SESs receiving RSG compared to RSG animals receiving bare metal stent or PESs. Quantitative polymerase chain reaction in human aortic endothelial cells (HAECs) revealed that sirolimus (but not paclitaxel) inhibited RSG-induced vascular endothelial growth factor transcription. Western blotting demonstrated that inhibition of molecular signaling in SES+RSG–treated arteries was similar to findings in HAECs treated with RSG and small interfering RNA to PPARγ, suggesting that sirolimus inhibits PPARγ. Transfection of HAECs with mTOR (mammalian target of rapamycin) short hairpin RNA and with Akt2 small interfering RNA significantly inhibited RSG-mediated transcriptional upregulation of heme oxygenase-1, a PPARγ target gene. Chromatin immunoprecipitation assay demonstrated sirolimus interferes with binding of PPARγ to its response elements in heme oxygenase-1 promoter.
Conclusions: mTOR/Akt2 is required for optimal PPARγ activation. Patients who receive SESs during concomitant RSG treatment may be at risk for delayed stent healing.
The local elution of cytostatic drugs at vessel wall sites with drug-eluting stents (DESs) reduces restenosis but has been associated with a long-term delay in healing (ie, lack of reendothelialization) and risk of late-stent thrombosis.1,2 Paclitaxel and sirolimus act via different mechanisms to reduce neointimal proliferation, yet it is unknown whether the difference in cellular targets of these 2 drugs is biologically or clinically relevant. Moreover, although local drug delivery minimizes systemic toxicity, it does not eliminate the potential for drug interaction between systemic medications and locally eluted drug. It is unknown whether this potential interaction could impact vascular healing.
Sirolimus (SRL) inhibits the mTOR (mammalian target of rapamycin), which plays an important role in connecting extracellular signals with intracellular pathways critical for arterial repair. mTOR may also have less well characterized effects on the regulation of gene transcription.3 Because considerable overlap exists between the molecular targets of SRL and agents used in the treatment of diabetes, the potential for drug interaction exists between systemically administered antidiabetic agents and locally eluted SRL. The thiazolidinedione (TZD) class of antidiabetic agents (which include rosiglitazone (RSG) and pioglitazone) are agonists of peroxisomal proliferator-activated receptor (PPAR)γ, a transcription factor expressed in smooth muscle and endothelial cells. TZDs are commonly prescribed to treat type 2 diabetes and have remarkable beneficial effects on insulin sensitivity; some studies suggest they also enhance endothelialization after arterial injury.4 Moreover, they are well known to have effects on the coronary vascular wall in doses normally given to humans.5 However, TZDs also have complex effects on many of the same signaling molecules affected by SRL and, as a result, might have potential interaction effects on vascular healing.
The mechanism of action of paclitaxel (Ptx) involves binding to the β subunit of the tubulin heterodimer, promoting tubulin polymerization, cell cycle arrest, and, eventually, inhibition of cell migration and proliferation.6 Although tubulin cycling is critical for cell division, it is independent of transmission of extra-/intracellular signaling. Thus, in contrast to the effects of SRL on the diverse actions of the mTOR pathway, the action of Ptx may make it independent of these signaling mechanisms.
To test our hypothesis that convergence of molecular signaling between oral antidiabetic medications such as the TZD PPARγ agonists and locally eluted SRL can result in significant effects on vascular healing, we examined the effect of the oral PPARγ agent rosiglitazone (RSG) in combination with sirolimus-eluting (SESs), Ptx-eluting (PESs), or bare-metal (BMSs) stents on endothelialization in the rabbit iliac model and explored the molecular mechanisms underlying the results.
An expanded Methods section providing details on cell culture, small interfering (si)RNA transfection, lentiviral transfection, Western blotting, quantitative polymerase chain reaction (PCR), and chromatin immunoprecipitation is available in the Online Data Supplement at http://circres.ahajournals.org.
Rabbit Model of Stent Implantation
Animals were stented as previously described.7 See the Online Data Supplement for more details.
Morphometric Analysis of Endothelial Surface Coverage
Composites of serial en face scanning electron microscopy images acquired at low power (×15 magnification). The extent of endothelial surface coverage above and between stent struts was conducted in a blinded fashion as previously described.7
Production of Vascular Endothelial Growth Factor From Stents Maintained in Organoid Culture
Select 14-day stents harvested for organoid culture were perfused in situ with Ringer’s lactate solution and processed as previously described.8 See the Online Data Supplement for more details.
A probability value of <0.05 was considered statistically significant. See the Online Data Supplement for more details.
All stents were successfully deployed without incidence of dissection or thrombosis. All animals remained healthy for the duration of the experiment. For 28-day scanning electron microscopy analysis, 1 artery in the SRL-RSG, Ptx-placebo, and Ptx-RSG groups was excluded because of processing artifact.
Endothelial Coverage Is Impaired by Rosiglitazone in Combination With SESs but Not BMSs or PESs
Overall, in terms of strut coverage and uncovered strut area in the placebo arm of each group, SESs demonstrated significantly less healing than PESs or BMSs. At 28 days after stent placement, endothelial coverage of stent struts between placebo- and RSG-treated animals was not significantly different in the BMS or PES groups as measured by scanning electron microscopy. However animals receiving SES+RSG had significantly less endothelial coverage of stent struts than animals receiving SES+placebo (Figure 1 and the Table).
Vascular Endothelial Growth Factor in the Arterial Wall Is Reduced When Rosiglitazone Is Combined With SESs but Increased When Rosiglitazone Is Combined With BMSs or PESs
Analysis of vascular endothelial growth factor (VEGF) protein levels in conditioned media collected from 14-day stented arterial segments (n=4 arteries per group) maintained in organoid culture for 48 hours showed a differential response to RSG treatment by stent type (Figure 2). Relative to placebo-treated animals (data expressed as percentages of control), VEGF levels in SES+RSG–treated animals were significantly reduced, whereas they were increased in animals treated with BMS+RSG and in animals treated with PESs. Pairwise group differences using these data with 95% confidence intervals are presented in Figure 2D, indicating that BMSs and PESs displayed similar levels of VEGF in the presence of RSG treatment.
Effects of Sirolimus, Paclitaxel, and Rosiglitazone on the Regulation of VEGF in Human Aortic Endothelial Cells
We explored whether RSG administered to endothelial cells in culture would upregulate VEGF transcription consistent with the findings in organ culture data. Human aortic endothelial cells (HAECs) were exposed for 24 hours to RSG in doses from 0.001 to 100 μmol/L. Although the minimum dose that significantly upregulated VEGF gene transcript was 50 μmol/L (1.6±0.3 fold, P=0.01), as detected by quantitative RT-PCR at 24 hours (Figure 3A), 100 μmol/L resulted in the most robust increase in VEGF transcription (6.5±1.5, P=0.0002) and was chosen for further in vitro studies. This dose is consistent with the upregulation of VEGF seen in our organ culture and in the blood of humans taking troglitazone, another TZD.9 Moreover, it is only slightly higher than the EC50 (10 μmol/L) of rosiglitazone for PPARγ activation in adipocytes.10
The dose of SRL used for subsequent studies is based on data from the rabbit iliac stent model in which local tissue concentrations averaged 1.5 μg per gram of tissue at 8 days after SES placement.11 This tissue concentration corresponds to 1.6 μmol/L/L at an assumed tissue density of 1g/cm. Because SRL tissue concentrations are known to decline thereafter, we chose a dose of 500 nmol/L because it is likely that in luminal wall cells tissue levels approximate this. The dose of Ptx for in vitro studies is based on similar data from the rabbit model where tissue levels averaged 1.3 μg per gram of tissue at 8 days after PES placement, which corresponds to 1.5 μmol/L.11 We chose 1.0 μmol/L for our in vitro work.
As seen in Figure 3B, RSG was observed to increase VEGF transcript in HAECs, which was significantly decreased by the addition of SRL to the media. SRL alone had no effect on VEGF transcription. These data suggests that one of the mechanisms by which SRL may reduce VEGF levels in animals treated with SES+RSG is through inhibition of RSG-induced transcriptional regulation of VEGF. Conversely, cotreatment of endothelial cells treated with RSG and Ptx had no effect on VEGF transcript (Figure 3C).
To explore the mechanism by which RSG upregulates VEGF, we used a siRNA targeted at PPARγ. HAECs transfected with PPARγ siRNA demonstrated significant inhibition of the level VEGF transcript when treated with RSG, compared to those receiving a scramble siRNA+RSG (Figure 3D). These results are consistent with a PPARγ-mediated mechanism of upregulation of VEGF and suggests that SRL interferes with this by blocking PPARγ-mediated transcription.
Effect of mTOR Inhibition by SRL on PPARγ-Mediated Gene Transcription
Although mTOR is best known for its role in regulation of mRNA translation, some data indicate it might have less well-characterized effects on transcriptional regulation. We examined the ability of SRL to regulate PPARγ activity by examining its effect on RSG-induced upregulation of a known PPARγ response element, heme oxygenase (HO)-1. Analysis of the human HO-1 promoter has revealed that it is transcriptionally regulated by PPARγ via 2 PPAR response elements.12 Figure 4A illustrates that HAECs treated for 24 hours with RSG demonstrated significant upregulation of HO-1 transcript, as measured by quantitative PCR. SRL alone and in combination with RSG significantly decreased but did not eliminate upregulation of HO-1 transcript by RSG. These data suggest that SRL may inhibit PPARγ-mediated gene regulation.
To determine whether mTOR is required for PPARγ activation, we transfected HAECs with recombinant lentivirus expressing either mTOR or scramble (control) short hairpin (sh)RNA and treated them with DMSO or RSG overnight. The upregulation of HO-1 transcript by RSG was significantly decreased in cells treated with mTOR shRNA compared to control (Figure 4B), indicating mTOR is required for PPARγ-mediated gene transcription.
To determine whether SRL inhibits PPARγ-mediated gene transcription by blocking its ability to bind to its PPAR response element in the promoter regions of DNA, we conducted a chromatin immunoprecipitation (ChIP) assay on endothelial cells exposed to DMSO, RSG, SRL, or RSG+SRL for 24 hours. We designed a positive primer encompassing the 2 PPAR response elements in the HO-1 gene and a “negative” primer that did not encompass this region (Online Table I). As can be seen in a representative gel (Figure 4C), both SRL alone and SRL+RSG demonstrated reduced PPARγ-mediated DNA binding as compared to DMSO or RSG alone. These data suggest that SRL inhibits PPARγ-mediated gene transcription by interfering with its ability to bind its response elements in the promoter regions of DNA.
Role of p70s6k Versus Akt in RSG-Mediated VEGF and HO-1 Transcription
To dissect the pathways critical in the upregulation of VEGF and HO-1 transcription by RSG, we used siRNAs targeted at the downstream effectors of both mTORC1 (ie, p70s6k) and mTORC2 (ie, Akt). HAECs were treated for 24 hours with DMSO versus RSG after transfection with scramble (control) versus p70s6k siRNA (Figure 4D). As expected, RSG significantly increased transcription of both VEGF and HO-1 as compared to DMSO-treated cells by quantitative PCR. However, p70s6k knockdown significantly inhibited the increase in VEGF transcription by RSG, whereas it had no effect of HO-1 transcription. It should also be noted that in DMSO-treated cells exposed to p70s6k knockdown alone had a borderline significant effect on VEGF transcription compared to control (P=0.09). These results indicate that p70s6k is not required HO-1 induction by RSG but is required for VEGF transcription in HAECs.
We also examined the role of Akt in these same processes using siRNAs. The Akt kinase family is composed of 3 highly homologous isoforms known from knockout mice experiments to function in the regulation of growth, glucose homeostasis, and neuronal development.13 In designing these experiments, we chose to target Akt2 because previous experiments in mice have shown knockdown produces a type 2 diabetes–like phenomenon with loss of adipose tissue.14 As seen in Figure 4D, as expected HAECs treated with RSG after transfection with control siRNA demonstrated significant increases in VEGF and HO-1 transcription compared to DMSO/control siRNA-treated cells. Akt2 knockdown significantly reduced HO-1 transcription in cells treated with DMSO or RSG versus those treated with DMSO/control siRNA. With regards to VEGF transcription, Akt2 knockdown significantly inhibited RSG-induced VEGF transcription and in fact significantly decreased VEGF transcription compared to cells treated with DMSO/control siRNA. These data indicate (1) that Akt2 is required for induction of HO-1 by RSG and (2) is also necessary for the upregulation of VEGF by RSG.
Effect of RSG on the mTOR Pathway
Next we examined the effects of SRL and RSG alone and in combination on the known targets of SRL to understand how RSG interacts with SRL in HAECs (Figure 5A). HAECs were treated for 24 hours in LSM supplemented with 25 ng/mL VEGF. As expected, SRL inhibited phosphorylation of mTOR and its downstream targets Akt, p70s6k, and 4EP-1 (data not shown). By densitometry, RSG had no effects on mTOR (data not shown) but increased Akt phosphorylation while simultaneously decreasing activation of p70s6k (Figure 5B). RSG did not affect 4E-binding protein (4E-BP)1 phosphorylation (data not shown). The combination of RSG and SRL resulted in similar reduction of mTOR, Akt, and 4E-BP1 phosphorylation as SRL alone but additive inhibition of phospho-p70s6k.
RSG Activates Akt Via PPARγ-Dependent Mechanisms but Regulates p70s6k Independent of Its Effects on PPARγ
To better understand the interaction of RSG and SRL, we examined the mechanisms by which RSG mediates Akt and p70s6k in HAECs. Although TZDs are synthetic ligands that activate PPARγ, some data suggest they have effects on protein synthesis independent of their ability to activate PPARγ.15 We explored the effects of RSG on p70s6k and Akt in HAECs using a siRNA to PPARγ. HAECs transfected with PPARγ siRNA demonstrated significant decreases in PPARγ protein levels as detected by Western blot compared to control-treated (scramble siRNA) cells 48 hours after transfection (data not shown). Figure 5C and 5D demonstrates that cells treated with RSG+scramble siRNA or RSG+PPARγ siRNA both demonstrated inhibition of phospho-p70s6k when compared to controls. These data demonstrate that RSG has inhibitory effects on protein translation independent of its ability to activate PPARγ. However, the effects of RSG on Akt were PPARγ-dependent because the increase in Akt phosphorylation by RSG/scramble siRNA cells was eliminated in cells treated with PPARγ siRNA+RSG.
We next examined these same signaling proteins using rabbit iliac arteries from animals treated with BMSs or SESs with and without RSG for 14 days. As seen in Figure 5C and 5D, Western blotting of arteries treated with BMS+RSG and SES+placebo demonstrated significant reduction in phosphorylated p70s6k when compared to animals treated with BMSs alone. Moreover, animals treated with SES+RSG demonstrated the most reduction in phospho-p70s6k. Similar to HAECs treated with RSG, phosphorylated Akt levels were higher than control in RSG-treated BMSs but significantly reduced to a similar extent in both SES+placebo and SES+RSG treated animals. These data mirror those using our PPARγ siRNA in endothelium (Figure 5B) and suggest that SRL blocks Akt phosphorylation by inhibiting PPARγ activation in vivo without affecting the effect of RSG on p70s6k.
Because p70s6k has been shown to regulate the translation and activity of hypoxia-inducible factor (HIF)-1α,16 a potent upregulator of VEGF, we also examined HIF-1α levels in these same samples. Figure 5C and 5D demonstrates that animals receiving SES+RSG had the least amount of HIF-1α, as detected by Western blot of 14 day-stented iliac artery samples.
Despite the relatively avid use of DESs to treat patients with diabetes mellitus and symptomatic coronary artery disease, to the best of our knowledge, this is the first published study to examine the effects of potential interactions between locally eluted drug and systemic medications used in the treatment of diabetes on vascular healing. We demonstrate for the first time that the mechanism of action by which SESs delay healing may have important clinical consequences because of convergent molecular signaling that takes place between mTOR and oral PPARγ agonists. Our results show that oral RSG significantly reduced endothelial strut coverage as compared to placebo in animals receiving SESs but not in animals receiving BMSs or PESs. Our data further suggest that mTOR/Akt2 activity is required for optimal PPARγ activation in vivo and that blockade of mTOR and its downstream mediators by SRL attenuates PPARγ activity. Moreover, commercially available PPARγ agonists may have inhibitory effects on mediators of protein translation independent of their ability to activate PPARγ. These interactions result in reduced endothelial coverage of stent struts, which may be associated with local decreases in VEGF expression in the arterial wall. No differences between rosiglitazone treatment and placebo arms were seen in animals receiving either PESs or BMSs. Based on our data, further study is warranted to determine whether patients receiving SESs and oral PPARγ agonists may be predisposed to increased risk of late-stent thrombosis secondary to lack of endothelial strut coverage. Moreover, the effect of other systemic oral pharmacological agents that share convergent molecular signaling with mTOR signaling on stent healing should be investigated.
Mechanisms of Delayed Healing
Although both SESs and PESs delay endothelialization of the stented segment compared to BMSs, their differing mechanisms of action result in substantial differences in this effect and their interaction with other signaling elements. Ptx is a lipophilic diterpenoid that binds to the β subunit of the tubulin heterodimer which promotes tubulin polymerization. Ptx suppresses endothelial cell proliferation and migration by disrupting microtubule dynamics, impacting cells in the mitosis phase of the cell cycle. Although Ptx has been shown to have effects on mitogen-activated protein kinase pathways, no evidence exists for interaction with mTOR signaling. Indeed we found that BMS- and PES-stented arteries behaved similarly to RSG in terms of VEGF upregulation. We also did not find any differences between PESs and BMSs in terms of percentage strut coverage at 28 days, consistent with previous data published by our group in this model.7 This finding is species-specific and should not be extrapolated to mean PESs and BMSs delay healing to a similar extent in humans. Many factors likely influence the healing process after PES placement, including direct inhibition of cellular proliferation and migration.17 Based on human pathological data, PESs and SESs may delay healing when compared to BMSs.1
Sirolimus is a specific inhibitor of the mTOR signaling pathway, known to play an important role in modulating cell division in response to mitogenic stimuli through control of protein synthesis. These actions are mediated though the formation of the mTORC1 complex which activates downstream effectors p70 ribosomal protein (p70s6k) and eukaryotic initiation factor 4E-BP1. In endothelial cells, sirolimus also prevents the formation of the mTORC2 complex (comprised of mTOR, rictor, Sin1, and mLST8), which has been shown to regulate Akt activity.18 Our results in HAECs confirm these observations and for the first time demonstrate that SESs inhibit arterial wall Akt activation in vivo.
Both mTORC1 and mTORC2 have been shown to play a role in processes critical to endothelial cell recovery. In various cell types, p70s6k activation itself has been shown to inhibit cell proliferation and survival, whereas both p70s6k and 4E-BP1 are involved in migration.19–21 mTORC2 regulates several Akt-related mechanisms including endothelial proliferation, migration, and survival.22 The overlapping roles of the downstream targets of mTORC1 and mTORC2 on processes related to vascular repair suggest that conditions or agents that share convergent molecular signaling with either pathway may have a significant impact on endothelial regrowth.
Although several lines of evidence have implicated the direct involvement of mTOR in the regulation of transcription, the mechanisms are not well understood. The only known target of mTOR is STAT3 (signal transducer and activator of transcription-3), which is reportedly phosphorylated and activated by mTOR.3 The work of Kim et al23 first suggested a role for mTOR in PPARγ activation in adipocytes, and our data confirm that mTOR is required for PPARγ activation in vascular endothelium, as evidence by the reduction in RSG induced HO-1 transcript with mTOR siRNA. Furthermore, dissection of downstream mTOR effectors p70s6k and Akt on HO-1 transcription using siRNAs indicate that Akt2, but not p70s6k, is required for PPARγ-mediated gene transcription (Figure 4). We have demonstrated for the first time that 1 mechanism by which mTOR inhibition decreases PPARγ-mediated gene transcription is through inhibition of DNA binding.
The exact mechanism by which Akt2 regulates PPARγ is not known and requires further investigation. It might inhibit the ability of PPARγ to bind to DNA through the regulation of other cofactors required in the PPAR/retinoid X receptor complex. One such cofactor is the winged helix transcription factor FOXO1 (forkhead box O-1), which is phosphorylated by mTORC2.18 When dephosphorylated, FOXO1 has been shown to inhibit PPARγ activity in part through disruption of DNA binding.24 In addition, it is also possible that mTOR may regulate other PPAR isoforms such as PPARα.
We also discovered that the synthetic PPARγ agonist RSG has inhibitory effects on p70s6k activity independent of its ability to activate PPARγ, resulting in profound inhibition of p70s6k activity in the arterial wall of SES+RSG animals, which was accompanied by reduction in arterial wall HIF-1α levels. Despite its effects on HIF-1α, RSG upregulated VEGF transcription and protein levels in HAECs and in the arterial wall, respectively, in a PPAR-dependent manner. These data suggest that RSG increases VEGF transcription independent of HIF-1α. Recently, Arany et al demonstrated that the transcriptional coactivator PPARγ coactivator (PGC)-1α regulates VEGF transcription independently of HIF-1 through coactivation of the orphan nuclear receptor estrogen related receptor α.25 PGC-1α can bind to and coactivate most members of the nuclear receptor family including PPARγ.26 RSG has been shown to upregulate PGC-1α transcription in human skeletal muscle and therefore it seems likely that the mechanism for increased transcription of VEGF may be through this pathway.27 Consistent with these findings, we found that knockdown of both p70s6k and Akt2 significantly inhibited RSG-induced VEGF transcription.
Based on our results, we have gained a greater understanding of PPARγ signaling mediators and their interactions with sirolimus, as show in Online Figure I. In the absence of sirolimus, activation of PPAR-responsive gene such as VEGF may lead to the activation of their respective receptors on the cell surface. This may result in a subsequent activation of the phosphatidyl inositol 3–Akt–mTOR pathway that is necessary for VEGF expression and reendothelialization of stents. In the presence of SRL, inhibition of Akt2 signaling, through inhibition of mTORC2 results in a reduction in VEGF expression and presumably a reduction in endothelial cell growth. Furthermore, our data demonstrates that mTOR is critical for PPARγ mediated gene transcription and SRL reduces PPARγ binding to its promoter, thereby reducing VEGF expression.
These findings raise interesting questions given the widespread use of SESs in patients with diabetes mellitus and the paucity of data regarding the differential vascular responses to SESs and PESs in diabetics. Although numerous clinical studies have demonstrated increased risk of late-stent thrombosis in diabetic patients receiving DESs, an understanding of the pathophysiology of this phenomenon is lacking.2 Moreover, no data exist regarding the long-term effect of SESs or PESs in patients treated with oral PPARγ agonists. It is also likely that the findings of this study also apply to the other commercially available PPARγ agonist pioglitazone. Our data suggest that patients with diabetes receiving these agents in combination with SESs should be followed more carefully. Although not specifically investigated in this study, newer DESs that elute mTOR inhibitors such as zotarolimus (Endeavor [Medtronic Vascular Inc]) and everolimus (Xience V [Abbott Laboratories]/PROMUS [Boston Scientific] stent) may produce similar results.
In addition, our data support the concept that convergent molecular signaling between locally eluted drug and systemic medications can have profound effects on vascular healing. The effect on healing of other conditions or agents commonly used by patients with symptomatic coronary disease that share convergent molecular signaling with mTOR should be investigated.
It should also be noted that BMSs with or without RSG demonstrated the least amount of uncovered strut area and emphasize that BMSs may represent an important alternative to DESs. This is supported by findings of the BASKET trial, which followed 826 patients randomized to DESs or BMSs for 6 months and found no significant differences in effectiveness between groups in patients with diabetes.28 Additional evidence from adequately powered randomized trials with diabetic patients is warranted.
The assessment of the effects of oral PPARγ agonists on vascular healing after DESs using a nonatherosclerotic, nondiabetic animal model may have underestimated the impact of oral PPARγ agonists in combination with DESs on vascular healing because both conditions are known to exacerbate many processes required for endothelial regrowth. Although animals models currently used in the assessment of stents are limited by their ability to replicate human conditions, results with the rabbit model have generally been representative of human responses.
This is the first study to examine the effects of oral PPARγ agonists in combination with commonly used DESs. Oral RSG significantly reduced endothelial strut coverage as compared to placebo in animals receiving SESs, but not in animals receiving BMSs or PESs. Our data suggest that mTOR/Akt2 is required for PPARγ activation and that RSG has inhibitory effects on p70s6k independent of its abilities to activate PPARγ. This convergence of signaling between mTOR inhibition and RSG results in reduced arterial wall VEGF production, an important mediator of reendothelialization after injury. Patients receiving oral PPARγ agonists in combination with SESs are potentially at increased risk for delayed healing.
We thank Lila Adams and Patricia Wilson (CVPath Institute Inc) for valuable technical assistance. We thank Dr David Harrison and Dr Martina Weber (Emory University) for assistance with the ChIP assay.
Sources of Funding
This study was supported in part by a grant from Boston Scientific. A.V.F. is supported by the Caryle Fraser Heart Center at Emory University.
A.V.F. has a sponsored research agreement with Boston Scientific, which supported part of this work. R.V. has received company-sponsored research support from Medtronic AVE, Abbott Vascular, W.L. Gore, Atrium Medical Corporation, Boston Scientific, NDC Cordis Corporation, Novartis, Orbus Medical Technologies, Biotronik, BioSensors, Alchimer, and Terumo and is a consultant for Medtronic AVE, Guidant, Abbott Laboratories, W.L. Gore, Terumo, and Volcano Therapeutics Inc. M.E. is an employee of Boston Scientific.
Original received May 6, 2009; revision received September 8, 2009; accepted September 10, 2009.
Daemen J, Wenaweser P, Tsuchida K, Abrecht L, Vaina S, Morger C, Kukreja N, Juni P, Sianos G, Hellige G, van Domburg RT, Hess OM, Boersma E, Meier B, Windecker S, Serruys PW. Early and late coronary stent thrombosis of sirolimus-eluting and paclitaxel-eluting stents in routine clinical practice: data from a large two-institutional cohort study. Lancet. 2007; 369: 667–678.
Hannan KM, Dilley RJ, de Dios ST, Little PJ. Troglitazone stimulates repair of the endothelium and inhibits neointimal formation in denuded rat aorta. Arterioscler Thromb Vasc Biol. 2003; 23: 762–768.
Choi D, Kim SK, Choi SH, Ko YG, Ahn CW, Jang Y, Lim SK, Lee HC, Cha BS. Preventative effects of rosiglitazone on restenosis after coronary stent implantation in patients with type 2 diabetes. Diabetes Care. 2004; 27: 2654–2660.
Emoto M, Anno T, Sato Y, Tanabe K, Okuya S, Tanizawa Y, Matsutani A, Oka Y. Troglitazone treatment increases plasma vascular endothelial growth factor in diabetic patients and its mRNA in 3T3-L1 adipocytes. Diabetes. 2001; 50: 1166–1170.
Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM, Kliewer SA. An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptor gamma (PPAR gamma). J Biol Chem. 1995; 270: 12953–12956.
Finn AV, Kolodgie FD, Harnek J, Guerrero LJ, Acampado E, Tefera K, Skorija K, Weber DK, Gold HK, Virmani R. Differential response of delayed healing and persistent inflammation at sites of overlapping sirolimus- or paclitaxel-eluting stents. Circulation. 2005; 112: 270–278.
Kronke G, Kadl A, Ikonomu E, Bluml S, Furnkranz A, Sarembock IJ, Bochkov VN, Exner M, Binder BR, Leitinger N. Expression of heme oxygenase-1 in human vascular cells is regulated by peroxisome proliferator-activated receptors. Arterioscler Thromb Vasc Biol. 2007; 27: 1276–1282.
Garofalo RS, Orena SJ, Rafidi K, Torchia AJ, Stock JL, Hildebrandt AL, Coskran T, Black SC, Brees DJ, Wicks JR, McNeish JD, Coleman KG. Severe diabetes, age-dependent loss of adipose tissue, and mild growth deficiency in mice lacking Akt2/PKB beta. J Clin Invest. 2003; 112: 197–208.
Cho DH, Choi YJ, Jo SA, Ryou J, Kim JY, Chung J, Jo I. Troglitazone acutely inhibits protein synthesis in endothelial cells via a novel mechanism involving protein phosphatase 2A-dependent p70 S6 kinase inhibition. Am J Physiol Cell Physiol. 2006; 291: C317–C326.
Hudson CC, Liu M, Chiang GG, Otterness DM, Loomis DC, Kaper F, Giaccia AJ, Abraham RT. Regulation of hypoxia-inducible factor 1alpha expression and function by the mammalian target of rapamycin. Mol Cell Biol. 2002; 22: 7004–7014.
Finn AV, Nakazawa G, Joner M, Kolodgie FD, Mont EK, Gold HK, Virmani R. Vascular responses to drug eluting stents: importance of delayed healing. Arterioscler Thromb Vasc Biol. 2007; 27: 1500–1510.
Dormond O, Madsen JC, Briscoe DM. The effects of mTOR-Akt interactions on anti-apoptotic signaling in vascular endothelial cells. J Biol Chem. 2007; 282: 23679–23686.
Vinals F, Chambard JC, Pouyssegur J. p70 S6 kinase-mediated protein synthesis is a critical step for vascular endothelial cell proliferation. J Biol Chem. 1999; 274: 26776–26782.
Harada H, Andersen JS, Mann M, Terada N, Korsmeyer SJ. p70S6 kinase signals cell survival as well as growth, inactivating the pro-apoptotic molecule BAD. Proc Natl Acad Sci U S A. 2001; 98: 9666–9670.
Shiojima I, Walsh K. Role of Akt signaling in vascular homeostasis and angiogenesis. Circ Res. 2002; 90: 1243–1250.
Kim JE, Chen J. Regulation of peroxisome proliferator-activated receptor-gamma activity by mammalian target of rapamycin and amino acids in adipogenesis. Diabetes. 2004; 53: 2748–2756.
Dowell P, Otto TC, Adi S, Lane MD. Convergence of peroxisome proliferator-activated receptor gamma and Foxo1 signaling pathways. J Biol Chem. 2003; 278: 45485–45491.
Mensink M, Hesselink MK, Russell AP, Schaart G, Sels JP, Schrauwen P. Improved skeletal muscle oxidative enzyme activity and restoration of PGC-1 alpha and PPAR beta/delta gene expression upon rosiglitazone treatment in obese patients with type 2 diabetes mellitus. Int J Obes (Lond). 2007; 31: 1302–1310.
Kaiser C, Brunner-La Rocca HP, Buser PT, Bonetti PO, Osswald S, Linka A, Bernheim A, Zutter A, Zellweger M, Grize L, Pfisterer ME. Incremental cost-effectiveness of drug-eluting stents compared with a third-generation bare-metal stent in a real-world setting: randomised Basel Stent Kosten Effektivitats Trial (BASKET). Lancet. 2005; 366: 921–929.