Superparamagnetic Iron Oxide–Based Method for Quantifying Recruitment of Monocytes to Mouse Atherosclerotic Lesions In Vivo
Enhancement by Tissue Necrosis Factor-α, Interleukin-1β, and Interferon-γ
Background— It has been found recently that the MRI contrast agent superparamagnetic iron oxide (SPIO) localizes to aortic atherosclerotic plaques. We therefore asked whether SPIO might be used to monitor monocyte recruitment into aortic atherosclerotic plaques.
Methods and Results— Eleven female apo E knockout (K/O) mice, each 11 months old, were divided into 2 groups. Six mice received tissue necrosis factor-α (0.2 μg IP once), interleukin-1β (0.2 μg IP once), and interferon-γ (100 U/g per day IP for 5 days); 5 received 0.5 mL saline containing1% BSA and served as sham-treated atherosclerotic controls. Two wild-type C57BL/6 mice served as sham-treated nonatherosclerotic controls. Three hours after initial cytokine or sham treatment, all mice received SPIO by intravenous injection (1 mmol/kg iron). Six days later, all mice were euthanized, the hearts and aortas were perfused under physiological pressure, and the entire aortas were studied histologically. Atherosclerotic plaques in cytokine-treated mice contained more iron-positive macrophages per cross section than did those in sham-treated apo E K/O control mice (42±11.8 versus 11.6±5.9) (P<0.0001). Iron-laden macrophages were present either in subendothelial plaque surfaces or in thin layers overlying the internal elastic lamina, often at the edges of atherosclerotic plaques. No iron deposition was seen in aortas of the wild-type nonatherosclerotic control mice. Immunocytochemistry showed mostly macrophages and few T lymphocytes in atherosclerotic plaques of cytokine-treated mice.
Conclusions— SPIO allows detection of iron-laden macrophages in the aortic subendothelium of apo E–deficient mice under basal conditions and monitoring of monocyte recruitment after cytokine injection.
Received October 11, 2002; revision received November 25, 2002; accepted December 9, 2002.
Atherosclerosis is an inflammatory disease.1 Inflammatory cells play a pivotal role in its initiation and progression. Increased macrophage density and/or activation in atherosclerotic plaque and the subsequently increased secretion of matrix metalloproteinases and possibly other proteases may induce the breakdown of collagen in the fibrous cap, thus contributing to the plaque’s vulnerability to rupture.2,3
Finding a noninvasive method for detecting monocyte recruitment into atherosclerotic plaques is a topic of great interest in atherosclerosis research.4–7 Such a technique could be used to assess plaque initiation, progression, and complications. Our group has been interested in the detection of inflammation in plaques by MRI 5 to 7 days after intravenous administration of superparamagnetic iron oxide (SPIO). SPIO is a nanoparticle that is avidly taken up by the reticuloendothelial system.8 In studies performed in Watanabe heritable hyperlipidemic rabbits (transgenic rabbit in which atherosclerosis develops spontaneously), Schmitz et al9 have shown that SPIO also homes to aortic atherosclerotic plaques, thus suggesting a potential noninvasive means for assessing atherosclerotic plaque.
Cytokines have long been known to enhance the recruitment of monocytes into atherosclerotic lesions.6,10,11 The purpose of the present study was to compare the uptake of iron (in the form of SPIO) into plaques of cytokine-treated apo E knockout (K/O) mice and age-matched, sham-treated control mice to (1) assess the effect of cytokines on monocyte recruitment and (2) validate SPIO as a marker of monocyte recruitment into atherosclerotic plaque.
Eleven female apo E K/O mice, each approximately 11 months old, were divided into 2 groups. The treatment group comprised 6 mice that received recombinant tissue necrosis factor-α (0.2 μg IP once), IL-1β (0.2 μg IP once), and interferon-γ (100 U/g IP daily for 5 days). The remaining 5 mice made up the sham-treated control group, which received 0.5 mL saline containing1% BSA. Two wild-type C57BL/6 mice were treated with saline and BSA and served as nonatherosclerotic controls. Three hours after initial cytokine or sham treatment, all animals were injected intravenously with SPIO (Feridex; Berlex Laboratories) (1 mmol/kg iron, undiluted, injected over 3 minutes).
All experimental procedures in these animals were performed in accordance with protocols approved by the Institutional Animal Care and Research Advisory Committee.
Histopathology and Immunohistochemistry
Six days after initial cytokine or sham treatment, recipient mice were euthanized with CO2, and the hearts and aortas were perfused under physiological pressure. In each case, the entire aorta from the sinuses of Valsalva up to the iliac bifurcation was formalin-fixed and serially sectioned transversely every 3 mm and stained with hematoxylin and eosin. Prussian blue and MAC-2 (Accurate Chemical) stains were used for detection of iron particles and macrophages, respectively.
The 6-day time point was chosen because work from our laboratory has shown that the highest MRI resolution is obtained 5 to 7 days after injection; corresponding histology also showed highest iron uptake around this time period.
A Student’s t test was used to test for statistical significance of the difference observed in number of iron-positive cells in cytokine-treated and sham-treated animals.
Histopathology and Immunohistochemistry
In the sham-treated apo E K/O control mice, iron-positive macrophages were deposited in the subendothelial surface of the atherosclerotic plaques (Figure 1, A and B). In the cytokine-treated mice, however, iron-positive macrophages were much more numerous and tended to localize in 2 places. In all 6 cases, the iron-laden macrophages localized in the subendothelium of plaques (Figure 1, C and D), where large, foamy cytoplasms were seen. In one section of two animals, they also penetrated deeper into the plaque overlying advanced and calcified plaques (Figure 1, C and D). In three mice, iron-laden macrophages localized in a thin layer over the internal elastic lamina, suggesting their recruitment into areas previously free of macrophage infiltration at the edges of atherosclerotic plaques (Figure 2). This phenomenon was seen in sham-treated apo E K/O control mice in a single case. In a single cytokine-treated mouse, iron-laden macrophages were also present in a single subendothelial layer in plaque-free areas (Figure 2C). Iron staining led to remarkable results; however, it was not uncommon to observe numerous brown golden particles in foamy macrophages on hematoxylin and eosin staining alone (Figure 2A). The endothelium overlying the plaques was unremarkable. No iron deposition was seen in the aortas of the wild-type control mice even after injection of cytokines.
Iron-positive cellularity was greater in cytokine-treated mice than in sham-treated control mice (42±11.8 versus 11.6±5.9 iron particles per cross section) (P<0.0001).
Immunohistochemistry showed that most cells in atherosclerotic plaques were positive for the macrophage stain Mac-2 and a few for the T-lymphocyte marker CD3 (Figure 3). One cytokine-treated mouse had a prominent mural infiltration of an intramural (muscular) coronary artery by foamy macrophages. The surrounding myocardium showed healing injury with mononuclear infiltration and early fine fibrosis (Figure 4).
The main finding of this study is that iron particles were detected in the subendothelium in aortic specimens from apo E K/O mice, and this effect was much greater in mice that received cytokines intraperitoneally and SPIO intravenously 6 days before they were killed. The iron was mostly present in atherosclerotic plaques, though most of the plaque area was free of iron because of the predominant superficial localization of the iron. There were, in addition, thin areas of iron-laden macrophages lying just above the internal elastic lamina, suggesting that these areas were free of monocyte infiltration before the administration of the cytokines and SPIO. Likewise, the cytokine-treated Apo E knockout mouse in which inflammation developed in an intramural coronary artery is presumably an example of a new lesion, since such lesions were not found in sham-treated Apo E knockout mice. Most of these zones were at the edge of atherosclerotic plaques. No iron deposition was seen in the aortas of wild-type nonatherosclerotic mice.
Histological correlation studies have shown that iron (as detected by Prussian blue staining) is not distributed homogeneously in atherosclerotic plaques but is mainly distributed in the subendothelium.9 Iron deposition was seen occasionally in endothelial cells (Figure 2C).
Although our present study was not designed to delineate the kinetics of the iron contained in SPIO particles, it is consistent with the evidence that most of the iron-laden macrophages in the cytokine-treated mice were newly recruited monocytes.6,12
The use of iron to study atherosclerotic plaques is not new. Iron, in the form of ferritin, was previously used by Gerrity4,13 in ultrastructural studies of atherosclerotic plaque to determine the phagocytic capacity of superficial macrophages. The author believed that the iron was taken up by the plaque macrophages from the circulating blood.
Three techniques have been reported for tracking the homing of macrophages into atherosclerotic plaques. The first, reported by Bylock and Gerrity,5 involves isolating swine monocytes from blood, labeling them with FITC, and reintroducing the labeled monocytes into the animal. Using this technique, Bylock and Gerrity showed that the monocytes attached themselves to thickened intima and atherosclerotic plaques. Also, Patel and colleagues10 used fluorescent-labeled macrophages to identify adhesion molecules involved in macrophages homing to plaque. The second technique, proposed by Steinberg et al7 and later perfected by the same group, involves transfusing monocytes from a male donor to a female recipient and tracking monocytes using the polymerase chain reaction (PCR) analysis of the aorta. Although highly informative, both techniques have drawbacks: (1) they cannot be used clinically, (2) they provide only a static picture of the interaction of monocytes with the arterial wall in atherogenesis, and (3) they leave the dynamics and control mechanisms of the sequence of events to be inferred from morphological data.
In contrast, a major advantage of the technique we described here, as shown in preliminary studies, is that SPIO is an MRI contrast agent that can be tracked noninvasively.8 Even though the dose used in the present study is ≈10 times those used in clinical medicine, SPIO was well tolerated clinically with no weight loss or other evidence of clinical toxicity. A second advantage is that histological techniques such as ours are better than PCR-based techniques at determining exactly where macrophages localize. An interesting finding in the present study was that in addition to homing in on plaques, some iron-laden macrophages also accumulated in nonplaque areas. In this respect, our study differed from that of Kim et al,6 in which the recruitment of monocytes to atherosclerotic lesions in LDL receptor–negative mice was quantified by PCR. Kim et al found that (1) lesion surface area was similar in control and cytokine-treated mice, (2) monocyte recruitment was enhanced by tissue necrosis factor-α and IL-1β administration, and (3) the response to cytokines was greater in younger mice with less advanced lesions than in older mice with more advanced lesions. Kim et al6 measured aortic lesion area with the naked eye by delineating the typical white opaque areas in the intima and quantified monocyte recruitment by PCR of DNA from pulverized aortic arch tissue.
Our finding that iron-laden macrophages were also present as single layers in the aortic subendothelial intima suggests that even when total plaque area may not appear to have increased macroscopically, initiation of new atherosclerotic plaques may already be apparent microscopically. Moreover, it is evident that the larger the gross plaque area, the smaller the uninvolved area, with less chance of single layers of iron-laden macrophages being deposited in grossly uninvolved areas. Thus, if one assumes that the lesion area had not increased in the study by Kim et al, it is possible that their findings can be explained in part by a recruitment of monocytes into grossly uninvolved areas that spuriously increased the number of macrophages recruited into plaques. However, since Kim et al6 studied their specimens 1 to 2 days after cytokine administration (and did not use interferon-γ) as opposed to our 6 days, it is possible that no single layer of macrophages was present that early. If, however, macrophages also accumulate on previously uninvolved areas after 1 to 2 days, the differences would be greater in cases with larger uninvolved areas (as in young mice), since the deposits in uninvolved areas would account for a higher percentage of newly recruited macrophages.
Although the present study was not designed to assess the topographic location of the newly recruited monocytes, we believe that a reasonable conclusion from our study in the face of the study by Kim et al6 is that newly recruited monocytes locate to the subendothelium into preformed plaques, at the edges of preformed plaques, and initiate new fatty streaks in areas previously free of lesions. Our study also does not rule out the possibility that a monocyte population without iron could have also been recruited into the plaque as well.
In summary, we have found that (1) SPIO can reveal the recruitment of monocytes into the subendothelial surface of already formed atherosclerotic plaques in cytokine-treated apo E K/O mice 6 days after inoculation and (2) SPIO homes in on new fatty streaks in the aorta.
This work was supported in part by the US Army’s Disaster Relief and Emergency Medical Services (DREAMS, DAMD 17–98–1-8002). The authors thank Selby Oberton, Dr Maziar Azadpour, and Dr Claudia E. Ferreira for help with the animal experiments.
Guest editor for this article is Prediman K. Shah, Cedars-Sinai Medical Center, Los Angeles, Calif.
Kim CJ, Khoo JC, Gillotte-Taylor K, et al. Polymerase chain reaction-based method for quantifying recruitment of monocytes to mouse atherosclerotic lesions in vivo: enhancement by tumor necrosis factor-alpha and interleukin-1 beta. Arterioscler Thromb Vasc Biol. 2000; 20: 1976–1982.
Steinberg D, Khoo JC, Glass CK, et al. A new approach to determining the rates of recruitment of circulating leukocytes into tissues: application to the measurement of leukocyte recruitment into atherosclerotic lesions. Proc Natl Acad Sci U S A. 1997; 94: 4040–4044.
Patel SS, Thiagarajan R, Willerson JT, et al. Inhibition of alpha4 integrin and ICAM-1 markedly attenuate macrophage homing to atherosclerotic plaques in ApoE-deficient mice. Circulation. 1998; 97: 75–81.
Rosenfeld ME. Leukocyte recruitment into developing atherosclerotic lesions: the complex interaction between multiple molecules keeps getting more complex. Arterioscler Thromb Vasc Biol. 2002; 22: 361–363.