Interferon-γ deficiency reduces neointimal formation in a model of endoluminal endothelial injury combined with atherogenic diet
- Authors:
- Published online on: June 15, 2012 https://doi.org/10.3892/ijmm.2012.1034
- Pages: 545-552
Abstract
Introduction
Restenosis is a common and serious complication associated with re-vascularisation by percutaneous coronary angioplasty, with considerable clinical implications (1). Inflammatory-mediated mechanisms following endothelial injury and/or stent placement are thought to play a pivotal role in restenosis (1). Moreover, injury-induced proliferation and/migration of vascular smooth muscle cells (VSMCs) is an important contributor to neointimal hyperplasia and in-stent restenosis (2,3). However, the precise underlying molecular mechanism(s) of restenosis are not entirely understood.
The pro-inflammatory cytokine interferon (IFN)-γ exhibits many biological functions including; mediation of inflammatory responses and immunomodulatory effects (4,5). IFN-γ is secreted from a variety of cells such as T helper cell type 1 (Th1) lymphocytes, activated macrophages, VSMCs and natural killer cells, and its expression has been demonstrated in atherosclerotic lesions (4,6). IFN-γ affects major cell types within the lesion and displays a complex role in vascular lesion development and progression (4). The biological effects of IFN-γ include; stimulating the differentiation of monocytes to macrophages, activation of macrophages and T-lymphocytes (4), and increasing the expression of platelet-derived growth factor (PDGF) and PDGF-β receptor on VSMCs, thereby mediating proliferation (7). In contrast, IFN-γ also exhibits protective effects against lesion formation; including inhibition of cultured VSMC proliferation (8,9), and induction of anti-inflammatory molecules; interleukin (IL)-1 receptor antagonist and IL-18 binding protein (10).
IFN-γ has been implicated in restenosis in both human and animal studies (11–13). Activation of IFN-γ signalling in neointimal SMCs was indicated in 37 of the 223 differentially expressed genes in atherectomy specimens of patients with restenosis, determined by a transcriptome study (13). Furthermore, elevated circulating IFN-γ has been detected within 15 min following stent implantation, in both chronic and acute coronary syndrome patients undergoing angioplasty, suggesting that modulation of IFN-γ release plays a role in inflammatory complications of angioplasty during/following stenting (12). Animal studies concerning the role of IFN-γ on injury-mediated models of neointimal formation have yielded variable results. Both administration of IFN-γ inhibitory protein (11), and exogenous IFN-γ treatment (14) reduced injury-induced neointimal formation in rats. Thus, the role of IFN-γ in injury-mediated neointimal formation is complex, and warrants further investigation.
Of note, a study demonstrated a pathogenic role for hypercholesterolemia in enhancing the degree of neointimal formation and VSMC proliferation, using a newly developed model of porcine vascular injury combined with a high cholesterol diet (15). Moreover, hypercholesterolemia induced systemic inflammation, with an increase in serum IFN-γ and TNF-α, and a local inflammatory response with higher T cell infiltrates (15).
Therefore, the aim of the current study was to determine the role of IFN-γ on injury-mediated neointimal formation, using a combination of a murine endoluminal endothelial-injury model with dietary-induced hypercholesterolemia. Our data demonstrated that IFN-γ−/− mice exhibit significantly reduced neointimal formation. These findings may provide useful insight for future design of therapeutic options for restenosis.
Methods and methods
Animals and diet
Male C57Bl/6 IFN-γ−/− (16) and WT mice were bred in the Blackburn Animal House, The University of Sydney. All experiments were performed in accordance with the guidelines of The University of Sydney Animal Care and Ethics Committee. Mice were housed under standard conditions, and allowed ad libitum access to water and standard mouse chow until ∼ 8 weeks of age. Subsequently, 1 week prior to probe-mediated endothelial injury of the iliac artery, mice were placed on an atherogenic diet containing 16% fat, 1% cholesterol and 0.5% cholate, obtained from Specialty Feeds (Cat. no. SF00-245, Glen Forrest, WA, Australia). The diet was continued after arterial injury until euthanisation.
Murine iliac artery endothelial injury model
Neointimal formation was induced in the left iliac artery via a standard procedure involving endoluminal mechanical disruption of the vessel wall (17,18). The resultant endothelial damage provides a primary stimulus for neointimal formation. Mice were anaesthetised via an intraperitoneal injection of Avertin, prior to surgery (19). Using a dissection microscope, the femoral artery, vein and nerve were identified and a custom-made probe (0.8-mm) with a roughened surface was passed via the femoral into the iliac artery. The insertion followed by withdrawing of the probe led to endothelial injury of the iliac artery. The femoral artery was then ligated proximal to the probe access point to stop the bleeding, but distal to the origin of the profunda femoris artery, allowing for continued blood flow through the iliac artery and collateral circulation to the limb. Following the surgical procedure, mice were allowed to recover in a sterile environment. The circulation and activity of the lower limb was monitored daily.
Plasma lipid analysis
Blood samples were collected via cardiac puncture at the time of sacrifice. Plasma was obtained and immediately frozen, and later analysed. Plasma TC and triglyceride levels were measured by specific enzymatic methods on a Cobas Modular autoanalyser (Roche Diagnostics Mannheim, Germany) and high-density lipoprotein cholesterol was measured using a direct enzymatic method on the same instrument. LDL-C was calculated using the Friedewald equation.
Excision of the iliac arteries
To evaluate the temporal development of neointima formation, iliac arteries were excised at 3 and 6 weeks post-surgery with each group consisting of five mice. The left (injured) and right (non-injured, control) iliac arteries were examined for the temporal development of neointima formation and normal intimal thickness, respectively. The iliac arteries and the surrounding muscle tissue were excised from the junction of the abdominal aorta to the ligated proximal segment of the artery, fixed overnight in 95% ethanol and embedded in paraffin wax as previously described (18).
Histopathology of the iliac arteries
Neointima formation was evaluated in 18 alternate serial sections (5 μm) taken through the centre third of the injured, and the corresponding non-injured (control) arteries. Sections were stained with hematoxylin and eosin (H&E), and photographed with an Olympus BX40 microscope attached to a DP71 digital camera (Olympus, Australia). The area of intima and media were evaluated using the computer image analysis software ImagePro Plus 4.5 (Diagnostic Instruments, USA), as previously described (18). Neointimal formation was quantified using two indices: i) %I/(I+M), and ii) %I/M. The counts from all sections of each artery were averaged to represent one mouse.
Immunohistochemistry
Representative paraffin sections (5 μm) were subjected to immunohistochemistry staining. Sections were incubated with rat anti-mouse F4/80 (WEHI, Melbourne, Australia) and CD31 (Abcam, Cambridge, UK) antibodies followed by a Horseradish Peroxidase (HRP)-conjugated rabbit anti-rat antibody (Dako, Sydney, Australia). Sections incubated with rabbit anti-mouse Ki-67 (Abcam) and TNF-α (Abcam) antibodies, and rabbit anti-human VWF (Dako), and PDGF-B (Abcam) antibodies with cross-reactivity with mouse VWF and PDGF-B, respectively, and further incubated with HRP-conjugated goat anti-rabbit antibody (Dako). Immunohistochemical staining for α-SMA and PCNA were performed using mouse anti-human antibodies that cross-react with mouse α-SMA (Dako) and PCNA (Abcam), respectively, and utilizing ARK™ (Animal Research kit) (Dako), according to the manufacturer’s instructions. All sections were then visualized with 3,3′ diaminobenzidine (DAB) (Dako), and lightly counterstained with hematoxylin.
Immunohistochemical quantitative evaluation
Immunohistochemical quantifications were performed on digitized images using image analysis software ImagePro Plus 4.5 (Diagnostic Instruments). Endothelialization was morphologically assessed in the immunohistochemical staining with anti-CD31 and anti-VWF antibodies. Re-endothelialization was calculated as the percentage of luminal surface covered by CD31 and VWF-positive cells to the total luminal surface. Proliferative activity was determined in the immunohistochemical staining with anti-PCNA and Ki-67 antibodies. The total number of brown-labelled positive nuclei within the neointima was counted, and a PCNA/Ki-67 labelling index was calculated [100% x (number of positive nuclei/total number of nuclei)]. The presence of monocytes/macrophages was assessed using anti-F4/80 antibody and the average number positive cells per cross section of the artery were determined. The neointimal α-SMA composition and TNF-α and PDGF-B expression were examined using image analysis software ImagePro Plus 4.5 (Diagnostic Instruments) to detect and express the positive (brown-labelled) area as a percentage of total neointimal area.
Statistical analysis
All data are expressed as means ± standard error of the mean (SEM). Data analysis was carried out by Student’s t-tests and two-way ANOVA using the statistical program in GraphPad Prism Version 4.0 (GraphPad Software, San Diego, CA, USA). A value of P<0.05 was considered statistically significant.
Results
Time-course of relative body weight and plasma lipid profile
All mice recovered well following surgery, and gained normal mobility and function within 24 h. The relative body weight of wild-type (WT) and IFN-γ−/− mice increased over the 7-week period, with no significant differences detected between the two groups (Fig. 1).
Placement of mice on the atherogenic diet for 7 weeks (corresponding to 6 weeks post-surgery), significantly elevated total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) levels in both WT and IFN-γ−/− mice, in comparison to each group’s corresponding standard diet group (P<0.05) (Table I). As expected, IFN-γ deficiency did not alter TC and LDL-C levels in the standard diet groups. However, in the atherogenic diet groups, IFN-γ deficiency resulted in ∼24 and 51% reduction in TC and LDL-C levels respectively, although without statistical significance.
Effect of IFN-γ deficiency on neointimal formation following endoluminal injury
The histology of the non-injured (control, right) iliac arteries of WT and IFN-γ−/− mice displayed three distinct layers: the intima, media and adventitia (Fig. 2). The arteries were of normal size and thickness, and no histological differences were detected between the two groups. The histopathology of the probe-injured (left) iliac arteries also displayed the three distinct layers (Fig. 2, weeks 3 and 6). However, in both groups, the intimal layer had a high cell density, was substantially thickened and temporal luminal narrowing was observed. A large proportion of the cells in the thickened intima are VSMCs, as confirmed by α-smooth muscle actin (SMA) immunohistochemical staining (Fig. 3).
The effect of IFN-γ deficiency on the extent of neointimal formation was evaluated from two indices using image analysis: i) the intimal area as a percentage of the intimal plus medial area [%I/(I+M)], and ii) the intimal area as a percentage of the medial area (%I/M) (Fig. 2). Both quantifications demonstrated temporal neointimal formation in the probe-injured iliac arteries of WT and IFN-γ−/− mice. From the 3-week time point, the WT group displayed near-occlusive neointimal formation. IFN-γ−/− mice exhibited significant delay in neointimal formation at the 3-week time point in comparison to their WT counterpart [%I/(I+M): WT, 56.2±2.5 vs. IFN-γ−/−, 41.1±3.4, P<0.01; %I/M: WT, 135.8±14.3 vs. IFN-γ−/−, 72.6±9.9, P<0.01]. By the 6-week time point, no significant differences were observed between the two groups.
The histological analysis of the non-injured (control, right) iliac arteries was carried out to examine baseline values for normal %I/(I+M) and %I/M for WT and IFN-γ−/− mice. Morphometry analysis of the non-injured arteries demonstrated consistent values across both time points, with no significant differences between the two groups [at 3 weeks post-surgery; %I/(I+M): WT, 10.8±0.4 vs. IFN-γ−/−, 10.9±0.6, P>0.05; %I/M: WT, 12.2±0.4 vs IFN-γ−/−, 12.3±0.8, P>0.05].
Detection of endothelium, neointimal PDGF-B expression, VSMC composition
The extent of endothelial cell coverage in the WT and IFN-γ−/− groups was assessed to examine whether differences in neointimal formation was due to differences in the extent of re-endothelialization following injury. A number of markers have been reported for endothelial cells; however many are additionally localized with other cell types (20). Therefore, to achieve a more reliable assessment of re-endothelialization, immunohistochemical staining was performed using both CD31 and von Willebrand factor (VWF) markers for endothelial cells. Near-complete endothelial cell coverage was detected in the injured arteries of WT and IFN-γ−/− mice, using both CD31 and VWF markers (Fig. 3). No significant differences in the extent of re-endothelialization were observed between the two groups, with CD31-positive cells covering 95.8±3.0 and 97.6±2.1%, and VWF-positive cells covering 92.4±2.9 and 85.6±2.9% (Fig. 3) of the arterial circumference at 3 weeks post-surgery in WT and IFN-γ−/− mice, respectively.
Immunohistochemical staining was performed for PDGF-B (Fig. 3), an important VSMC mitogen and chemoattractant (21). PDGF-B was detected in both intima and media layers. IFN-γ−/− mice exhibited reduced neointimal proportion of PDGF-B-positive area in comparison to their WT counterpart, with a significant difference at 3 weeks post-surgery (WT, 8.6±1.9% vs. IFN-γ−/−, 2.5±0.4%; P<0.05) (Fig. 3). To examine neointima composition, the presence of VSMCs (α-SMA) (Fig. 3) was examined by immunohistochemistry. Image analysis revealed that VSMCs are a major component of the neointima in this model, consistent with our previous findings (17,18). Similarly, IFN-γ−/− mice exhibited significantly reduced proportion of neointimal VSMCs in comparison to their WT counterpart, with a significant difference at 3 weeks post-surgery (WT, 40.6±0.8% vs. IFN-γ−/−, 22.5±1.5%; P<0.05) (Fig. 3).
Macrophage infiltration, TNF-α expression, and proliferative activity
Immunohistochemical analysis was performed to examine macrophages/monocytes infiltration (F4/80) (Fig. 4) within the vessel. Macrophages were primarily detected in the adventitia layer of the arteries and only a few were present within the neointima, consistent with our previous finding (17). IFN-γ−/− mice displayed a significant reduction in the number of vessel macrophages in comparison to their WT counterpart at both 3 (WT, 24.5±3.1% vs. IFN-γ−/−, 10±0.5%; P<0.05) and 6 weeks post-surgery (WT, 17±2.3% vs. IFN-γ−/−, 6.5±1.8%; P<0.01) (Fig. 4). Deficiency in both IFN-γ and the pro-inflammatory cytokine TNF-α has been shown to synergistically inhibit injury-mediated neointimal formation (22), and therefore TNF-α expression was assessed by immunohistochemistry (Fig. 4). IFN-γ−/− mice displayed lower neointimal TNF-α expression at both time points, although without statistical significance (Fig. 4).
To assess whether the differences in neointimal formation are due to variations in cellular proliferation, immunohistochemistry using the cell proliferation markers; Ki67 and PCNA was performed. Although Ki-67 nuclear antigen is expressed in G1 through M phase and absent from G0 phase, the antibody may not recognise all proliferating cells, such as those with a long, or at the end of, the G1 phase or those entering the S phase of the cell cycle (23). However, proliferating nuclear cell antigen (PCNA) signal increases in G1 phase, with a maximum at S phase, and declines during G2/M phase of the cell cycle (24). Therefore, both PCNA and Ki-67 labelling were assessed to take into account these differences (Fig. 4). Higher Ki-67 labelling index was observed in comparison to PCNA, in both groups of mice. IFN-γ−/− mice exhibited a significant reduction in neointimal labelling index for both PCNA (WT, 4.0±0.5% vs. IFN-γ−/−, 2.5±0.4%; P<0.05) and Ki-67 (WT, 82.1±1.6% vs. IFN-γ−/−, 75.0±2.0%; P<0.05) (Fig. 4), in comparison to their WT counterpart, at 3 weeks post-surgery.
Discussion
The present study demonstrates that in this model of endoluminal injury combined with dietary-induced hypercholesterolemia, the overall effect of IFN-γ is to enhance neointimal formation. IFN-γ deficiency led to a significant reduction in endothelial injury-induced neointimal formation at the 3-week time point. In WT mice, endoluminal injury of the iliac artery resulted in rapid neointimal responses at the initial time point (3 weeks) and appeared to stabilize by the final time point (6 weeks). IFN-γ−/− mice exhibited delayed injury-induced neointimal formation at the earlier time point, although achieved a similar extent of neointimal formation by the final time-point, in comparison to their WT counterpart. The significant reduction in neointimal formation in the IFN-γ−/− group is consistent with decreased macrophage infiltration, and reduced neointimal VSMC composition, PDGF-B expression and cellular proliferation, with no differences in re-endothelialization between the two groups of mice.
Restenosis is multifactorial, and has been attributed to elastic recoil, neointimal proliferation induced by vessel wall injury, and negative remodelling (1). Inflammatory mechanisms have been indicated to play a pivotal role in neointimal formation and restenosis. The pleiotropic pro-inflammatory cytokine IFN-γ has been implicated in the pathogenic processes of restenosis (11–13). However, IFN-γ has been shown to both promote (11) and inhibit (14) injury-mediated neointimal formation, and therefore warrants further investigation.
Of note, mild hypercholesterolemia has been demonstrated to enhance the degree of stenosis and VSMC proliferation, and significantly elevate systemic IFN-γ and TNF-α levels, in a newly developed model of porcine vascular injury combined with a high cholesterol diet (15). Additionally, hypercholesterolemia has been shown to induce macrophage infiltration following balloon injury via NF-κB activation and increase ICAM-1 expression in rabbits (25), and increase PDGF-A and -B mRNA expression in circulating mononuclear cells in patients, in comparison to normocholesterolemic individuals (26). The effect of hypercholesterolemic state on the role of IFN-γ in the acute injury model remains to be explored. Therefore, the current study utilised an injury-induced model of restenosis combined with the atherogenic diet to examine the role of IFN-γ on neointimal formation in a hypercholesterolemic state, in an attempt to recapitulate the processes that occur in humans following angioplasty. Notably, the Th1 pattern of cellular immune responses is predominant in human atheroma and in animal studies of atherosclerosis (27). Thus, in the present study C57Bl/6 strain of mice was utilized, as this strain is thought to be amongst the most Th1-responsive strains, and therefore more responsive to deficiency in Th1 polarising cytokines, principally IFN-γ (28).
In the present study, the inclusion of the atherogenic diet led to elevated TC and LDL-C levels, as previously documented (29), whilst IFN-γ deficiency did not significantly alter TC and LDL-C levels, in line with the findings of non-injury atherosclerotic model in LDLR−/− mice (30), and ApoE−/− mice (28). Furthermore, the inclusion of the atherogenic diet accelerated neointimal formation in WT mice, in comparison to our previous study of the endoluminal injury model in normocholesterolemic WT mice (17), indicating that dietary-induced hypercholesterolemia promotes injury-mediated neointimal formation.
Endothelial injury is an important contributing mechanism to restenosis, and endothelial regeneration has inhibitory effects on neointimal formation and development of restenosis (31). In the present study, re-endothelialization was near complete in both WT and IFN-γ−/− mice at the 3-week time point, indicating that IFN-γ deficiency did not alter the extent of re-endothelialization, consistent with the findings in BALB/c mice on standard diet (22). Such data suggest that the differences in neointimal formation between WT and IFN-γ−/− mice are due to injury-induced neointimal responses, rather than variations in regulation of endothelial growth.
Macrophage infiltration plays an important role in the pathogenesis of restenosis, supported by a higher abundance in atherectomy tissue from patients that develop restenosis (32). Moreover, systemic inactivation and depletion of monocytes/macrophages results in reduced neointimal formation in injury models (33). In the present study, the significant reduction in vessel macrophage infiltration detected in the IFN-γ−/− mice is in line with the important role of IFN-γ in activation and/or recruitment of monocytes/macrophages (4,5). Thus, it is speculated that the decreased macrophage infiltration in the IFN-γ−/− mice may have altered the pool of vessel cytokines and growth factors (34), attenuating injury-mediated vessel responses and leading to reduced neointimal formation in these mice. In this study, macrophages were predominantly found within the adventitia layer of the artery, consistent with our previous finding (17). This is in support of the increasing evidence that vascular inflammation begins in the adventitia and progresses inward towards the intimal layer (35).
In addition, PDGF-mediated pathway may be responsible for the observed reduction in neointimal formation in the IFN-γ−/− mice. PDGF is produced by a number of cells including activated macrophages and SMCs, and PDGF-BB has a well-recognised role in the pathogenesis of restenosis (36). In the present study, reduced infiltrating macrophage in the IFN-γ−/− group may have contributed to lower PDGF-B chain expression, corresponding to decreased neointimal formation at the 3-week time point. Moreover, PDGF-BB gene expression has been demonstrated to enhance neointimal formation by stimulating proliferation and migration of SMC into the intima layer, as well as increasing extracellular matrix synthesis (21). Furthermore, exogenous IFN-γ has been shown to potentiate the proliferative effect of PDGF-BB and up-regulate PDGF-β receptor in VSMCs, without having direct mitogenic effects (7). Thus, in the current study, the reduced PDGF-B chain expression may be responsible for the significant reduction in neointimal VSMCs and proliferation in the IFN-γ−/− mice, consistent with lower neointimal formation, suggesting that the PDGF-B-mediated pathway plays an important role in neointimal formation in this model.
Within the pro-inflammatory cytokine cascade, there is redundancy and overlap of effects, therefore interpretation of data obtained in relation to gene deficiency is complex (37). In the present study, IFN-γ−/− mice reached the same stable neointimal size as the WT group at the 6-week time point, despite the IFN-γ−/− group displaying reduced macrophage infiltration. This may be due to inflammatory pathway redundancy or other factors influencing injury-mediated neointimal responses. Interestingly, deficiency in both IFN-γ and TNF-α has been shown to synergistically inhibit neointimal formation (22). In the present study, IFN-γ−/− mice displayed reduced neointimal TNF-α expression, although without statistical significance. Further evaluation of other cytokines and growth factors may allow for clarification of this multifaceted biological process.
There are a number of limitations associated with this study. The model used in the present study is injury-mediated; however, the procedure was performed on non-atherosclerotic arteries, and may not accurately resemble the vascular processes that occur in human coronary artery angioplasty. Moreover, whilst the inclusion of the atherogenic diet allowed for induction of hypercholesterolemia in a timely manner, the effects of the diet may not precisely represent human hypercholesterolemia. Thus, further investigations are required to accurately determine the role of IFN-γ in the pathogenesis of restenosis in clinical settings.
In conclusion, these data demonstrated that the absence of IFN-γ markedly reduced neointimal formation, consistent with reduced inflammatory responses and proliferative effects, in a mouse model of endoluminal injury combined with dietary-induced hypercholesterolemia. However, in the absence of IFN-γ, other inflammatory pathway(s) may exhibit compensatory effects with time. This suggests that in combination, modulating the IFN-γ-mediated pathway as well as targeting other inflammatory cytokines and growth factors may prove to be beneficial in controlling neointimal formation and restenosis.
Acknowledgements
We thank Dr Jane Radford for her expert technical immunohistochemical assistance.
