Both Apolipoprotein E and Immune Deficiency Exacerbate Neointimal Hyperplasia After Vascular Injury in Mice
In this study, we investigated the role of T and B lymphocytes in neointimal hyperplasia after endothelial denudation. Catheter-induced endothelial denudation of wild-type mice resulted in rapid infiltration of lymphocytes to the site of injury. Mice defective in recombination-activating gene 2 (RAG2−/−) showed increased neointimal formation 14 days after vascular injury in comparison to their wild-type immune-competent littermates. Immunohistochemical studies revealed the preponderance of smooth muscle cells and a significantly higher number of proliferating cells in the neointima of the RAG2−/− mice. The neointima size and the number of proliferating smooth muscle cells in the injured vessel of RAG2−/− mice were similar to those observed in the injured arteries of apolipoprotein E (apoE)-deficient (apoE−/−) mice. Interestingly, mice with double apoE and RAG2 gene mutations (apoE−/− RAG2−/−) displayed similar neointimal characteristics as mice with a single gene defect, suggesting a similar mechanism for apoE and lymphocyte protection against injury-induced neointimal formation. The protective role of lymphocytes against neointimal formation after vascular injury directly contrasts to their reported role in the promotion of atherosclerosis, which was observed in both apoE+/+ and apoE−/− mice. Thus, these results support the hypothesis of different etiology between hyperlipidemia-induced atherosclerosis and injury-induced vascular occlusion.
One of the early events in the pathogenesis of atherosclerosis is the infiltration of macrophages and lymphocytes to the lesion area.1–3⇓⇓ Results from these histological examinations have led to an explosion of studies that focused on the role of the immune response in atherogenesis. Most of these studies have used mouse models to determine how the lack of specific genes involved in the inflammatory response pathway may affect the development of the atherosclerotic plaque. For example, the inhibition of T-cell activation via suppression of CD40 signaling was shown to reduce the severity of atherosclerosis.4 Mice with T-cell deficiency were also shown to have reduced atherosclerotic plaque in response to an atherogenic high-fat diet.5 Interestingly, significantly lower aortic root atherosclerosis was observed in the immune-incompetent apolipoprotein E (apoE)-null (apoE−/−) mice than in immune-competent apoE−/− mice only under low-fat dietary conditions.6,7⇓ Immune competency was not a factor in brachiocephalic trunk atherosclerosis under these conditions.7 However, lymphocyte-deficient apoE−/− mice showed aortic root atherosclerosis similar to those present in immune-competent apoE−/− mice when they were both fed a Western type high-fat diet.6–8⇓⇓ Taken together, these results suggest interactive effects between dietary lipids, apoE, and immune competency in dictating the severity of atherosclerosis, especially in the aortic root. This hypothesis was further supported by additional studies with immune-incompetent LDL receptor (LDLr)-deficient mice. In the LDLR−/− background instead of apoE−/− background, T-cell deficiency was found to decrease early progression of atherosclerosis when the animals were fed the high-fat diet.9
The effect of lymphocytes on another form of vascular occlusive disease, such as those induced as a consequence of catheter- or balloon-mediated endothelial denudation, is less clear. Studies with athymic nude rats and T lymphocyte-depleted rats yielded equivocal results.10,11⇓ Additionally, although interferon-γ secreted by T lymphocytes was shown to inhibit smooth muscle cell proliferation in vitro and it limits arterial proliferative lesions in response to arterial injury,12 cyclosporine suppression of lymphocyte activation had no effect on injury-induced neointimal proliferation in cholesterol-clamped rabbits.13 However, cyclosporine may have effects in addition to its suppression of lymphocyte activation. Despite certain inconsistencies in experimental results, these studies clearly indicated differences in the impact of lymphocytes in diet-induced atherosclerosis and endothelial denudation-induced stenosis of the arterial wall.
In addition to the immune cell system, apoE level is also an important determinant of neointimal formation after vascular injury in mice.14 Transgenic overexpression of apoE is protective against neointimal hyperplasia whereas apoE gene deletion exacerbates neointima formation.14 The mechanism by which apoE protects against injury-induced neointimal hyperplasia may be the result of its ability to suppress smooth muscle cell migration and proliferation in response to growth factors.15–17⇓⇓ However, apoE has also been shown to have immunosuppressive properties, and it limits mitogen-induced lymphocyte activation.18,19⇓ Thus, apoE may protect vascular response to injury by multiple mechanisms. The present study compares apoE−/− mice, RAG2−/− mice, and apoE−/−RAG2−/− double knockout mice for their susceptibility to neointimal hyperplasia after vascular injury. The goal is to assess the potential interactive effects of apoE and lymphocytes in modulating vascular response to injury.
The apoE−/− mice that had previously been backcrossed with C57BL/6 mice for 10 generations were obtained from Jackson Laboratories (Bar Harbor, Me). The RAG2−/− mice, which are defective in expression of the RAG2 thus resulting in T and B lymphocyte deficiency,20 were backcrossed with C57BL/6 mice for 4 generations before use. The apoE−/−RAG2−/− double knockout mice were generated by crossing apoE−/− and RAG2−/− mice as described.7 Because of possible genetic differences in vascular response to injury,14 both wild-type C57BL/6 and [C57BL/6×129] mice were used as controls. The animals were maintained in a specific pathogen-free environment on a 12-hour light/12-hour dark cycle and were fed a normal mouse chow diet (Harlan Teklad). Food and water were available ad libitum. Male animals at 6 to 8 weeks of age, weighing 25 to 30 g were used in this study. All animal experimentation protocols were carried out under the guidelines of animal welfare by the University of Cincinnati, in accordance with National Institutes of Health guidelines.
Mouse Carotid Artery Injury
Mechanically induced carotid artery injury was performed as described elsewhere.14,21⇓ Briefly, animals were anesthetized with a mixture of ketamine (80 mg/kg, Fort Dodge Laboratories) and xylazine (16 mg/kg, The Butler Co). The left external carotid artery was looped proximally and ligated distally with 7-0 silk sutures (Ethicon). An epon-resin probe made by forming a bead slightly larger than the diameter of the carotid artery (approximately 0.45 mm) on a 3-0 nylon suture was used for the arterial injury. A transverse arteriotomy was made between the 7-0 sutures, and the resin probe was inserted. The probe was advanced toward the aortic arch and withdrawn 5 times. The probe was then removed, and the proximal 7-0 suture was ligated. Animals were allowed to wake up under a warming lamp. The identical surgical procedure was applied to each animal to assure reproducibility of the results. This injury procedure results in consistent complete endothelial denudation, as demonstrated by both Evans Blue staining21 and immunohistochemical analysis with antibodies to von Willebrand factor,14 with minimal trauma to the underlying internal elastic lamina. The occasional animals displaying damaged elastic lamina after this injury procedure were excluded from subsequent characterization.
Tissue Preparation, Histology, and Morphometry
Mice were killed 1 hour or 14 days after injury by perfusion fixation with 10% phosphate-buffered formalin (pH 7.0) for 15 minutes at physiological pressure. The whole-neck that contains both uninjured right and injured left common carotid arteries was harvested and fixed with the same solution for an additional 48 hours. The whole-neck was decalcified before being embedded in paraffin. Ten identical 5-μm cross-sections at 500-μm intervals were made from the distal side of the whole-neck beginning at the point of the proximally ligated 7-0 suture. Parallel sections were subjected to routine hematoxylin and eosin staining as well as to Verhoeff Van-Gieson staining of elastic lamina. Unstained sections from each level were used for immunohistochemistry.
Morphometric analyses were performed with Verhoeff Van-Gieson–stained sections. For each animal, four whole-neck cross-sections were digitized, and the vessels were quantified with Scion Image analysis software (Scion). Measurements of luminal area, area inside the internal elastic lamina (IELA), and the area encircled by external elastic lamina were made on each section. Medial area was calculated by subtracting IELA from external elastic lamina, and intimal area (IA) was calculated by subtracting luminal area from IELA. From these measurements, the ratio of IA and medial area (I/M ratio) was calculated. Luminal stenosis was represented as IA/IELA×100%. For each animal, results from the four levels of the whole-neck section were averaged.
Identification of Proliferating Cells
Cells in the S-phase of the growth cycle were identified based on immunohistochemical staining with antibodies against 5-bromo-2‘-deoxyuridine (BrdU) after its injection into the mice. The mice were injected intraperitoneally with three 50-mg/kg doses of BrdU (Sigma Chemical Co) at 24, 8, and 1 hour before the were killed. Formalin-fixed, paraffin-embedded consecutive sections were used for immunohistochemical staining with mouse antibody against BrdU (see below). The number of proliferating cells (BrdU-positive cells) was determined in both medial and intimal areas of arterial cross-sections. Measurements were made in a blinded manner.
Monoclonal anti-α smooth muscle actin antibody (Clone 1A4, diluted 1:3000, Sigma Chemical Co) and monoclonal anti-BrdU antibody (Clone BU 33, diluted 1:300, Sigma Chemical Co) were used as described previously to identify smooth muscle cells and proliferating cells in the arterial wall.14 Acetone-fixed frozen sections were used for immunohistochemical identification of macrophages and lymphocytes. Macrophages were identified by staining with monoclonal anti-mouse CD11b (Clone M1/70, diluted 1:50, BD PharMingen). Lymphocytes were identified by reactivity with anti-mouse CD4 or anti-mouse CD8 (both diluted 1:100, San Cruz Biotechnology). Splenic tissues from wild type and RAG2−/− mice were used as positive and negative controls, respectively. Positive staining was detected by reaction with avidin-biotin complex reagent (Vectastain Elite ABC kit, Vector Laboratories) and aminoethyl carbazole (AEC, Zymed Laboratories). Sections were counterstained with hematoxylin.
All values were expressed as mean±SEM. When only two groups (injured arteries and contralateral control arteries) were compared, differences were assessed by using a paired Student’s t test. Multiple comparisons were first tested by using ANOVA. When the ANOVA demonstrated significant differences, individual mean differences were analyzed with the Student-Newman-Keuls test. Statistical software SigmaStat (Version 2.0, Jandel Co) was used in statistical analysis. For all statistical analyses, P<0.05 was considered significant.
Consistent with results reported earlier,14 catheter-induced carotid arterial injury in apoE−/− mice resulted in massive neointimal formation detectable after 14 days (Figure 1). In comparison, both wild-type C57BL6 mice and [C57BL/6×129] mice were resistant to injury-induced neointimal formation (Figure 1). Immunohistochemical staining of the neointima in apoE−/− mice with smooth muscle–specific α-actin antibodies identified smooth muscle cells as the major cell type present in the neointima (Figure I, please see http://atvb.ahajournals.org).
The potential involvement of immune cells in arterial response to injury was explored by determining leukocyte infiltration into the injured arteries 1 hour after the induction of endothelial denudation. Lymphocytes were found to adhere to the denuded vessel wall of wild-type C57BL/6 and apoE−/− mice, as judged by immunohistochemical staining with antibodies against CD4 and CD8 (Figure 2). No lymphocyte was found along the injured arterial wall of the RAG2−/− mice (Figure 2). Interestingly, no CD4- or CD8-positive cells nor macrophages were found in the vessel wall 14 days after vascular injury when neointima was apparent (Figure II, please see http://atvb.ahajournals.org).
The role of lymphocytes in neointimal hyperplasia after arterial injury was explored by comparing apoE−/−, RAG2−/−, and apoE−/−RAG2−/− mice, along with their respective controls, for neointimal formation in response to endothelial denudation. Representative vessel sections of uninjured and injured cartoid arteries from these animals are shown in Figure 1 and morphometric analysis of the data are shown in Figure 3. The results showed that the medial area of the injured arteries increased significantly 14 days after injury in comparison to their respective controls for all strains of mice except [C57BL/6×129] mice, but they are not significantly different from one another (Figure 3). The injured arteries of apoE−/− mice also displayed a significantly enlarged neointima in comparison to that observed in the injured arteries of C57BL/6 mice (12181.76±3104.17 versus 2990.58±1353.49 μm2, P<0.05, Figure 3B). The RAG2−/− mice also showed marked increases of neointimal formation after arterial injury compared with their wild-type [C57BL/6×129] mice (18112.31±7111.43 versus 1202.09±92.69 μm2, P<0.05, Figure 3B). As in the neointima of the apoE−/− mice, the neointima in the RAG2−/− was predominantly smooth muscle cells (Figure I). Interestingly, the neointima observed in the injured arteries of the RAG2−/− mice was very similar to that observed in the apoE−/− mice (Figure 3).
To determine if apoE deficiency and RAG2 deficiency act through similar or different mechanisms in promoting neointimal formation in response to arterial injury, endothelial denudation of the carotid arteries in apoE−/−RAG2−/− double knockout mice was performed. Results showed that the neointima developed in these animals 14 days after arterial injury was similar to those observed in mice with a single gene defect of either apoE or RAG2 (13574.96±2524.03 μm2, P=not significant versus apo E−/− and RAG2−/− mice, respectively. Figure 3B). The I/M ratio in the injured arteries from these strains of mice (Figure 3C) demonstrated similar results with no significant differences observed in the injured arteries of apoE−/−, RAG2−/−, and apoE−/−RAG2−/− mice. The I/M ratio in these genetically modified mice was all significantly higher than that observed in the C57BL/6 and [C57BL/6×129] wild-type mice (Figure 3C). The increased neointimal area in the injured arteries of apoE−/−, RAG2−/−, and apoE−/−RAG2−/− mice also resulted in significant increase of luminal stenosis in these animals (Figure 3D). Although luminal stenosis appeared to be higher in some apoE−/−RAG2−/− mice compared with that in apoE−/− and RAG2−/− animals (for example, see Figure 1), analysis of data collected from the entire study did not show statistically significant difference in luminal stenosis between these groups (Figure 3).
The exacerbated neointimal formation observed in the injured arteries of apoE−/− mice has been attributed to the failure in suppression of smooth muscle cell proliferation in the absence of this apolipoprotein.14,15⇓ Deficiencies in interferon-γ production, caused by the lack of T lymphocytes, may also explain the increased neointimal size in the RAG2−/− mice after vascular injury.10,12⇓ The next experiment was designed to determine smooth muscle cell proliferation index in the injured arteries of apoE−/−RAG2−/− mice and to compare this characteristic with that observed in apoE−/− and RAG2−/− mice and in their respective wild-type controls. Cell replication was quantitated by the number of BrdU-labeled cells per cross-section in both the media and the neointima at 14 days after vascular injury. Because the majority of cells within the neointima are smooth muscle cells, the number of BrdU-labeled cells reflects the smooth muscle cell proliferation index. Consistent with results reported previously,14 the total number of BrdU-labeled cells was greatly increased in the neointimal area of the injured arteries of apoE−/− mice in comparison with their control C57BL/6 wild-type mice (Figure III, please see http://atvb.ahajournals.org). The injured arteries of both RAG2−/− and apoE−/−RAG2−/− mice also displayed significantly higher level of BrdU-positive cells in comparison to those present in the injured arteries of their wild-type controls (Figure III). The number of proliferating cells present in the neointima of apoE−/−, RAG2−/−, and apoE−/−RAG2−/− mice was not significantly different (Figure 4). Interestingly, the number of BrdU-positive cells in the media was highest in mice with apoE deficiency, but medial cell proliferation in the RAG2−/− mice was not statistically different from that in the wild-type controls (Figure 4).
Results of the present study revealed lymphocyte infiltration as one of the earliest events after vascular injury, with both CD4- and CD8-positive cells adhering to the vessel wall within 1 hour of endothelial denudation. This process was transient with no immune cells detectable at the injured site after 14 days. Our results also suggested that the early lymphocyte infiltration is a protective event against massive smooth muscle cell hyperplasia in the neointima. The latter conclusion was derived from the observation of exacerbated neointimal formation and increased number of proliferating smooth muscle cells in the injured arteries of the RAG2−/− mice. The severity of neointimal hyperplasia, as observed by BrdU-labeling of smooth muscle cells, and vascular occlusion, as measured by luminal stenosis, were similar to those observed in the injured arteries of apoE−/− mice. Interestingly, immune-incompetent apoE−/− mice also showed similar neointimal hyperplasia and vascular occlusion after vascular injury, suggesting that lymphocytes and apoE may serve similar and/or overlapping role in protecting vascular occlusion in response to endothelial denudation.
Although the precise mechanism by which apoE inhibits vascular response to injury has not been determined conclusively, recent in vitro data suggest that apoE may protect against injury-induced vascular occlusion through multiple mechanisms. Firstly, apoE is capable of suppressing agonist-induced platelet aggregation and activation.22,23⇓ Second, apoE has been shown to limit inflammatory response by suppressing lymphocyte activation in response to mitogenic stimulation.19,24,25⇓⇓ Finally, apoE has also been shown to directly inhibit smooth muscle cell migration and proliferation.15,16⇓ It is not clear which if any of these putative actions of apoE is dominant in the in vivo situation. The observation of exacerbated neotinimal formation in immune-incompetent mice expressing normal amounts of apoE argues against the immunoregulatory function of apoE as the mechanism for its vascular protective effect after endothelial denudation. Our data does not exclude the other mentioned effects of apoE in inhibiting neointimal formation after vascular injury. Inhibition of platelet activation and/or suppression of smooth muscle cell migration and proliferation remain as possible influences of apoE in its suppression of neointimal formation.
The role of immune cells in vascular response to injury has been controversial. Our data with the RAG2−/− mice support the hypothesis that lymphocytes play a protective role in preventing exacerbated neointimal hyperplasia after endothelial denudation. The most striking observation reported here is that exacerbated neointimal formation is seen in the face of deficiency of either apoE or T and B cells or both deficiencies together. The neointima is quantitatively similar in each case, and the two genetic interventions do not produce an additive effect on neointimal formation. This suggests that apoE and immune cells or their products are components of a common pathway though their precise mechanism of action may or may not be identical. ApoE may be involved in the recruitment and activation of lymphocytes, whose products such as interferon-γ may then influence neointimal formation. Alternatively, both apoE and lymphocytes may be components of a common pathway in which they act in sequence. We do not know which component is upstream or downstream in this pathway. Both components influence the migration and/or proliferation of smooth muscle cells though the precise step of their involvement in this process may differ.
It is possible that apoE influences platelet activation and release of platelet-derived growth factor. It also may act on the transduction of the platelet-derived growth factor signaling to smooth muscle cell proliferation or migration. The major product of T cells interferon-γ has been shown to inhibit smooth muscle cell proliferation though its detailed biochemical mechanism of this effect is not clear.26
The results of this study revealed that lymphocyte depletion increases neointimal hyperplasia after vascular injury, and also suggested that early infiltration of CD4- and CD8-positive lymphocytes may be protective against injury-induced neointimal hyperplasia. These results are consistent with those reported by other investigators.10,12⇓ The protective role of lymphocytes against injury-induced neointimal hyperplasia is in contrast to their effects on atherosclerosis as seen in the aortic root.6,7,9⇓⇓ In fact, CD4-positive T cells have been shown to aggravate atherosclerosis in the apoE−/− mice.27 These results add to the current literature suggesting differences in the etiology of vascular occlusive disease between injury-induced vasculopathy and diet-induced atherosclerosis.28 Previously, we have already shown that C57BL/6 mice, which are more susceptible to diet-induced atherosclerosis in comparison with the FVB/N mice,29 were resistant to injury-induced neointimal formation whereas the FVB/N mice were found to be susceptible.14 However, this should not be taken to indicate that all features of injury-induced vasculopathy and diet-induced atherosclerosis are different. Atherosclerosis is a complex process, which involves, among other cellular processes, smooth muscle cell migration and proliferation. This latter process is involved in vascular neointimal formation also. Thus, vascular stenosis and atherosclerosis may share some components of the processes though the overall processes exhibit significant differences. These differences may not lie in the control of smooth muscle cell dynamics. However, the fact that the atherosclerosis-sensitive strain C57BL/6 is relatively resistant to neointimal formation tends to argue against this formulation.
In devising specific therapeutic approaches to atherosclerosis and vascular stenosis, it is important that there is an awareness of the common features as well as of the significant differences. This is so that the treatment of one does not compound the other. To obtain this facility, it is thus of critical importance that we achieve greater mechanistic understanding of the smooth muscle cell dynamics in atherosclerosis and vascular stenosis.
An observation that may bear on a part of the mechanism of neointimal formation is worthy of comment. All vessels were perfusion-fixed under physiological pressure. However, the arteries in apoE−/− mice, whether injured or not, and injured RAG2−/− vessels showed elastic laminae that are not extended but recoiled or wavy. Because endothelial cell–derived nitric oxide is important in vasorelaxation,30,31⇓ the apparent vasoconstriction observed in the injured vessels of the genetically modified mice may be the result of the loss of endothelial NO synthase as a consequence of endothelial denudation. Although vascular injury of control C57BL/6 and [C57BL/6×129] mice also resulted in endothelial denudation, the normal elastic laminae observed in these vessels suggests that endothelial denudation alone cannot account for the vasoconstrictive phenotype observed in the injured vessels of the apoE−/− and RAG2−/− mice. It is possible that nitric oxide produced by inducible NO synthase expressed in smooth muscle cells32 and/or by the recruited T lymphocytes33 after arterial injury may compensate for the lack of endothelial NO synthase in producing sufficient nitric oxide to maintain vasorelaxation. Nitric oxide synthesized in smooth muscle cells can also inhibit their proliferation and reduce neointimal hyperplasia.32 However, in the absence of apoE, a protein that has been shown to promote nitric oxide production in a variety of cell types,16,22,34⇓⇓ the reduced capacity for nitric oxide synthesis may account for the contracted state and the exacerbated neointima in the vessel wall of the apoE−/− mice. The apparent vasoconstriction observed in the uninjured vessel wall of apoE−/− mice is supportive of the importance of apoE in maintaining vascular homeostasis. Lymphocytes adhered to the denuded vessel wall may also secrete interferon-γ that can stimulate nitric oxide production by smooth muscle cells.35 Thus, the lack of lymphocytes in the RAG2−/− mice may compromise this compensatory mechanism, resulting in nitric oxide deficiency and the observed vasoconstriction and exacerbated neointima in the injured vessels of these animals. These latter possibilities need to be explored further.
This research was supported by National Institutes of Health grants HL61332 (to D.Y.H.) and HL56827 (to G.S.G.). We gratefully acknowledge the helpful discussions, technical advice and assistance of David Kuhel, April Netus, and Sherry Crabtree during the course of this study.
Received December 12, 2001; revision accepted January 9, 2002.
- ↵Dansky HM, Charlton SA, Harper MM, Smith JD. T and B lymphocytes play a minor role in atherosclerotic plaque formation in the apolipoprotein E–deficient mouse. Proc Natl Acad Sci U. S. A. 1997; 94: 4642–4646.
- ↵Reardon CA, Blachowicz L, White T, Cabana V, Wang Y, Lukens J, Bluestone J, Getz GS. Effect of immune deficiency on lipoproteins and atherosclerosis in male apolipoprotein E–deficient mice. Arterioscler Thromb Vasc Biol. 2001; 21: 1011–1016.
- ↵Hansson GK, Holm J, Holm S, Fotev Z, Hedrich HJ, Fingerle J. T lymphocytes inhibit the vascular response to injury. Proc Natl Acad Sci U. S. A.. 1991; 88: 10530–10534.
- ↵Hansson GK, Holm J. Interferon-gamma inhibits arterial stenosis after injury. Circulation. 1991; 84: 1266–1272.
- ↵Andersen HO, Hansen BF, Holm P, Stender S, Nordestgaard BG. Effect of cyclosporine on arterial balloon injury lesions in cholesterol-clamped rabbits: T lymphocyte–mediated immune responses not involved in balloon injury–induced neointimal proliferation. Arterioscler Thromb Vasc Biol. 1999; 19: 1687–1694.
- ↵Ishigami M, Swertfeger DK, Granholm NA, Hui DY. Apolipoprotein E inhibits platelet-derived growth factor–induced vascular smooth muscle cell migration and proliferation by suppressing signal transduction and preventing cell entry to G1 phase. J Biol Chem. 1998; 273: 20156–20161.
- ↵Ishigami M, Swertfeger DK, Hui MS, Granholm NA, Hui DY. Apolipoprotein E inhibition of vascular smooth muscle cell proliferation but not the inhibition of migration is mediated through activation of inducible nitric oxide synthase. Arterioscler Thromb Vasc Biol. 2000; 20: 1020–1026.
- ↵Swertfeger DK, Hui DY. Apolipoprotein E receptor binding versus heparan sulfate proteoglycan binding in its regulation of smooth muscle cell migration and proliferation. J. Biol Chem. 2001; 276: 25043–25048.
- ↵Curtiss LK. ApoE in atherosclerosis: a protein with multiple hats. Arterioscler Thromb Vasc Biol. 2000; 20: 1852–1853.
- ↵Hui DY, Harmony JAK, Innerarity TL, Mahley RW. Immunoregulatory plasma lipoproteins: role of apoprotein E and B. J Biol Chem. 1980; 255: 11775–11781.
- ↵Riddell DR, Graham A, Owen JS. Apolipoprotein E inhibits platelet aggregation through the l. arginine:nitric oxide pathway: implications for vascular disease. J Biol Chem. 1997; 272: 89–95.
- ↵Curtiss LK, Edgington TS. Immunoregulatory plasma low density lipoprotein: the biologic activity and receptor binding specificity is independent of neutral lipids. J Immunol. 1981; 126: 1008–1012.
- ↵Pepe MG, Curtiss LK. Apolipoprotein E is a biologically active constituent of the normal immunoregulatory lipoprotein, LDL-In. J Immunol. 1986; 136: 3716–3723.
- ↵Hansson GK, Hellstrand M, Rymo L, Rubbia L, Gabbiani G. Interferon-gamma inhibits both proliferation and expression of differentiation-specific alpha-smooth muscle actin in arterial smooth muscle cells. J Exp Med. 1989; 170: 1595–1608.
- ↵Zhou X, Nicoletti A, Elhage R, Hansson GK. Transfer of CD4+ T cells aggravates atherosclerosis in immunodeficient apolipoprotein E knockout mice. Circulation. 2000; 102: 2919–2922.
- ↵Ishida BY, Paigen B. Atherosclerosis in the mouse. Monogr Hum Genetics. 1989; 12: 189–222.
- ↵Ignarro LJ, Buga GM, Wood KS, Byrns RE, Chaudhuri G. Endothelium-derived relaxing factor produced and released from artery and vein is nitric oxide. Proc Natl Acad Sci U. S. A. 1987; 84: 9265–9269.
- ↵Yan ZQ, Hansson GK. Overexpression of inducible nitric oxide synthase by neointimal smooth muscle cells. Circ Res. 1998; 82: 21–29.
- ↵Geng Y, Hansson GK, Holme E. Interferon-gamma and tumor necrosis factor synergize to induce nitric oxide production and inhibit mitochondrial respiration in vascular smooth muscle cells. Circ Res. 1992; 71: 1268–1276.