Level of Macrophage uPA Expression Is an Important Determinant of Atherosclerotic Lesion Growth in Apoe−/− Mice
Objective— Enhanced plasminogen activation, mediated by overexpression of urokinase-type plasminogen activator (uPA), accelerates atherosclerosis in apolipoprotein E–null mice. However, the mechanisms through which uPA acts remain unclear. In addition, although elevated uPA expression can accelerate murine atherosclerosis, there is not yet any evidence that decreased uPA expression would retard atherosclerosis.
Methods and Results— We used a bone marrow transplant (BMT) approach and apolipoprotein E–deficient (Apoe−/−) mice to investigate cellular mechanisms of uPA-accelerated atherosclerosis, aortic dilation, and sudden death. We also used BMT to determine whether postnatal loss of uPA expression in macrophages retards atherosclerosis. BMT from uPA-overexpressing mice yielded recipients with macrophage-specific uPA overexpression; whereas BMT from uPA knockout mice yielded recipients with macrophage-specific loss of uPA expression. Recipients of uPA-overexpressing BM acquired all the vascular phenotypes (accelerated atherosclerosis, aortic medial destruction and dilation, severe coronary stenoses) as well as the sudden death phenotype of uPA-overexpressing mice. Moreover, fat-fed 37-week-old recipients of uPA-null BM had significantly less atherosclerosis than recipients of uPA wild-type marrow (40% less aortic surface lesion area; P=0.03).
Conclusions— The level of uPA expression by macrophages—over a broad range—is an important determinant of atherosclerotic lesion growth in Apoe−/− mice.
Blood-derived cells of the monocyte-macrophage lineage play a central role in the initiation and progression of atherosclerosis.1 However, the detailed molecular mechanisms through which macrophages contribute to lesion growth remain incompletely defined. Identification of these mechanisms would yield a better understanding of atherosclerotic plaque initiation and progression and might identify new approaches to the prevention of atherosclerosis and its complications.
Urokinase plasminogen activator (uPA) is a serine protease that is expressed by macrophages within human atherosclerotic plaques2,3 and therefore may play a significant role in both early and advanced atherosclerosis. A role for macrophage-expressed uPA in atherosclerosis is suggested by: (1) a positive correlation between artery wall uPA expression and the severity of human atherosclerosis4; (2) human studies that associate increased plasminogen activation with accelerated atherosclerosis5–7; and (3) well-established roles for uPA—and its physiological substrate plasminogen—in processes that are central to the development of atherosclerosis and its complications, including cell migration, matrix and growth factor metabolism, and inflammation.8,9 Based on this theoretical framework, several years ago we generated apolipoprotein E–deficient (Apoe−/−) mice that express a macrophage-targeted uPA transgene and used these mice to test the hypothesis that increased macrophage uPA expression would accelerate atherosclerosis. These uPA-overexpressing (“SR-uPA”) mice had increased aortic atherosclerosis, severe coronary artery stenoses, and premature sudden death, all in a plasminogen-dependent manner.10,11 However, because expression of the uPA transgene was not clearly restricted to macrophages (transgene mRNA was present in several organs including liver, heart, and aorta),10 it remained possible that “leaky” uPA transgene expression in another cell type was responsible for accelerated atherosclerosis or the other phenotypes of the SR-uPA mice. Moreover, our finding of increased atherosclerosis in these uPA-overexpressing mice contrasts with two studies that compared atherosclerosis in Plau−/−Apoe−/− and Plau+/+Apoe−/− mice.12,13 Both of these studies reported that complete loss of uPA expression (in Plau−/−Apoe−/−mice) did not affect aortic root atherosclerosis.
To test more rigorously whether accelerated atherosclerosis in SR-uPA mice is attributable to expression of uPA by macrophages and to determine whether loss of macrophage-expressed uPA would affect atherosclerosis, we used bone marrow transplant (BMT) approaches to augment or delete uPA expression specifically in macrophages of Apoe−/− mice. Our results support our initial hypothesis that elevated uPA expression by arterial wall macrophages accelerates atherosclerosis. We also provide the first experimental evidence that loss of uPA expression—specifically in macrophages—retards atherosclerotic plaque growth.
An expanded Methods section is available online at http://atvb.ahajournals.org. We used two experimental approaches: (1) Lethal irradiation of Apoe−/− mice and bone marrow transplantation (BMT) from transgenic uPA-overexpressing or control nontransgenic donors (all uPA wild-type, Apoe−/−); (2) Lethal irradiation of Apoe−/− mice and BMT from uPA-null (Plau−/−) or control Plau+/+ donors (all Apoe−/−). Key techniques and lines of mice are described in previous studies or adapted from them, including: mice with macrophage-targeted overexpression of uPA (SR-uPA+/0 mice),10 Plau−/− mice,14 lethal irradiation and bone marrow transplantation,15,16 measurement of plasminogen activator activity in medium conditioned by peritoneal macrophages, bone marrow–derived macrophages, and explanted aortae,10,11 plasma lipid analyses,17,18 morphological, histological, and immunohistochemical analyses of aortic root atherosclerotic lesions, coronary stenoses, and aortic surface atherosclerosis,10,11 in vitro macrophage chemotactic migration,19 in vitro macrophage activation,20,21 and statistical methods.10,11
Successful Hematopoietic Reconstitution After Bone Marrow Transplantation
PCR amplification of peripheral blood DNA from female recipients of male SRuPA+/0Plau+/+, SR-uPA0/0Plau−/−, or SR-uPA0/0Plau+/+ BM revealed abundant donor-derived Y chromosomal DNA in all cases (not shown). Our previous FACS analyses of identically irradiated mice reconstituted with green fluorescent protein-expressing BM revealed approximately 80% green-fluorescent peripheral blood cells.22
Recipients of SR-uPA+/0Plau+/+ BM Have Elevated PA Activity in Macrophages and Aortae Whereas Plau−/− Macrophages Have No PA Activity
Medium conditioned by peritoneal macrophages of SR-uPA+/0Plau+/+ BM recipients contained far more PA activity than medium conditioned by macrophages of SR-uPA0/0 Plau+/+ BM recipients (650 [570–920] versus 0.43 [0.36 to 0.62] IU/108 cells/h; P=0.04; Figure 1A). Similarly, medium conditioned by aortae from recipients of SR-uPA+/0Plau+/+ BM contained more PA activity than medium conditioned by aortae from recipients of SR-uPA0/0Plau+/+ BM (7.2 [6.7 to 21] versus 1.5 [1.3 to 1.7] IU/mg protein; P=0.04; Figure 1B). As discussed in our initial report,10 uPA activity is increased far less in SR-uPA+/0 aortae than in macrophage-conditioned medium. The magnitude of increased uPA activity in SR-uPA+/0 aortae is within the range reported for atherosclerotic versus normal human arteries.4
PA activity of peritoneal macrophages and aortae of recipients of SR-uPA0/0Plau+/+ BM was low, near the limit of detection of the assays. Therefore, we did not harvest peritoneal macrophages and aortae from recipients of SR-uPA0/0Plau−/− BM to determine whether their PA activity was less than PA activity of peritoneal macrophages and aortae of recipients of SR-uPA0/0Plau+/+ BM. Instead, we measured PA activity of BM-derived macrophages from nontransplanted SR-uPA0/0Plau+/+ and SR-uPA0/0Plau−/− mice. Medium conditioned by BM-derived SR-uPA0/0 Plau+/+ macrophages contained a low level of PA activity (2.9 [2.3 to 11] IU/1010 cells/h); whereas, no PA activity was detectable in medium conditioned by SR-uPA0/0Plau−/− macrophages (<0.8 IU/1010 cells/h; n=4 each; P=0.03).
Negligible Expression of uPA mRNA in Peripheral Blood Leukocytes of SR-uPA+/0Plau+/+ and SR-uPA0/0Plau+/+ Mice
To test whether transplantation of SR-uPA+/0Plau+/+ BM would increase expression of uPA by BM-derived cells other than macrophages, we used quantitative RT-PCR to measure uPA mRNA in peripheral blood leukocytes from 4 SR-uPA+/0 Plau+/+ mice. Peripheral blood leukocyte uPA mRNA was near or below the limit of detection in all 4 mice, with GAPDH mRNA easily detected in all samples (data not shown). Moreover, in peripheral leukocyte subpopulations of SR-uPA+/0Plau+/+ mice, uPA mRNA was undetectable in 3 of 3 samples of lymphocytes, 3 of 4 samples of monocytes, and 1 of 2 samples of granulocytes. When detected, uPA mRNA was at near-background levels (supplemental Figure I and data not shown). We conclude that: (1) there is no significant expression of the SR-uPA transgene in peripheral blood leukocytes; and (2) recipients of SR-uPA+/0Plau+/+ BM do not acquire uPA-overexpressing peripheral blood leukocytes.
Similarly, to test whether transplantation of SR-uPA0/0 Plau−/− BM to SR-uPA0/0Plau+/+ mice would decrease uPA expression by BM-derived cells other than macrophages, we performed quantitative RT-PCR on peripheral blood leukocyte RNA of 2 SR-uPA0/0Plau+/+ mice. No uPA mRNA was detected (data not shown). Therefore, transplantation of SR-uPA0/0Plau−/− BM cannot decrease uPA expression in peripheral blood leukocytes because SR-uPA0/0 Plau+/+ leukocytes do not express uPA. Because neither the SR-uPA transgene nor the endogenous uPA gene are expressed by peripheral blood leukocytes: (1) transplantation of either SR-uPA+/0Plau+/+ or SR-uPA0/0Plau−/− BM to SR-uPA0/0Plau+/+ recipients does not alter uPA expression by peripheral blood leukocytes; and (2) phenotypes found in SR-uPA0/0Plau+/+ recipients of SR-uPA+/0 Plau+/+ or SR-uPA0/0Plau−/− BM cannot be attributed to altered uPA expression in peripheral blood leukocytes. Instead—with the caveat that we did not formally exclude differentiation of transplanted SR-uPA+/0 BM precursors to nonhematopoietic cells such as endothelial cells—phenotypes in recipients of SR-uPA+/0Plau+/+ BM must be attributable to uPA overexpression by SR-uPA+/0Plau+/+ macrophages10 and phenotypes in recipients of SR-uPA0/0 Plau−/− BM must be attributable to deficient uPA expression by Plau−/− macrophages. Stated differently, SR-uPA0/0Plau+/+ recipients of SR-uPA+/0Plau+/+ BM likely have macrophage-specific uPA overexpression and SR-uPA0/0Plau+/+ recipients of SR-uPA0/0Plau−/− BM are essentially macrophage-specific uPA-knockout mice.
Reconstitution With SR-uPA+/0Plau+/+ Marrow Increases Plasma uPA but Reconstitution With SR-uPA0/0Plau−/− Marrow Does Not Decrease Plasma uPA
Because uPA is a secreted protein and uPA-secreting macrophages accumulate in aortae and hearts of SR-uPA+/0 mice,10,22 we tested whether plasma uPA was elevated in SR-uPA+/0Plau+/+ mice or recipients of SR-uPA+/0Plau+/+ BM. We measured plasma uPA antigen in 10- to 15-week-old BM donor mice (SR-uPA+/0Plau+/+ as well as SR-uPA0/0 Plau+/+ and SR-uPA0/0Plau−/−) and at the time of harvest in all 6 groups of BM recipients (Figure 2). Plasma uPA antigen was significantly higher in SR-uPA+/0 Plau+/+ donors than in SR-uPA0/0Plau+/+ donors (100±11 versus 2.7±0.42 ng/mL; Figure 2A; P<0.01). As expected, uPA antigen was undetectable in plasma of SR-uPA0/0Plau−/− mice (<0.05 ng/mL; P<0.01 versus SR-uPA0/0Plau+/+ mice). Plasma uPA antigen in SR-uPA0/0Plau+/+ recipients of SR-uPA+/0 Plau+/+ marrow increased after BMT to levels similar to those of SR-uPA+/0Plau+/+ donors (140±16 ng/mL; Figure 2B; P=0.1 versus SR-uPA+/0Plau+/+ donors). In contrast, plasma uPA antigen in SR-uPA0/0Plau+/+ recipients of SR-uPA0/0Plau−/− BM did not decrease (3.5±0.18 versus 3.6±0.42 ng/mL for recipients of SR-uPA0/0Plau+/+ 29 weeks after BMT; P=0.8). Seven weeks after BMT, recipients of SR-uPA0/0Plau−/− BM had a small increase in plasma uPA (4.9±0.37 versus 3.5±0.31 ng/mL for recipients of SR-uPA0/0Plau+/+ BM; P=0.02).
Reconstitution With SR-uPA+/0Plau+/+or SR-uPA0/0Plau−/− Bone Marrow Does Not Affect Plasma Lipids or Peripheral Blood Monocyte Counts
Plasma cholesterol and triglycerides did not differ between recipients of SR-uPA+/0Plau+/+ BM and recipients of SR-uPA0/0Plau+/+ BM (Table 1). Plasma cholesterol and triglycerides also did not differ between recipients of SR-uPA0/0 Plau−/− BM and recipients of SR-uPA0/0Plau+/+ BM at 37 weeks of age (fat-fed recipients; Table 2). As expected, plasma cholesterol and triglyceride levels of chow-fed recipients of SR-uPA0/0Plau−/− or SR-uPA0/0Plau+/+ BM were lower than in fat-fed mice, but they were equivalent between the 2 groups of recipients (Table 2). Plasma FPLC profiles did not differ between recipients of SR-uPA+/0Plau+/+ or SR-uPA0/0Plau+/+ BM, between 37-week-old fat-fed recipients of SR-uPA0/0Plau−/− or SR-uPA0/0Plau+/+ BM, or between chow-fed 15-week-old recipients of SR-uPA0/0 Plau−/− or SR-uPA0/0Plau+/+ BM (supplemental Figure II). At both 15 and 37 weeks of age, peripheral blood monocyte counts did not differ between recipients of SR-uPA0/0Plau−/− BM and recipients of SR-uPA0/0Plau+/+ BM (both fat-fed and chow-fed recipients; Table 2). We reported previously that monocyte counts of SR-uPA+/0Plau+/+ mice do not differ from those of SR-uPA0/0Plau+/+ mice.10
Transplantation of SR-uPA+/0Plau+/+ Bone Marrow Causes Early Mortality, Accelerated Atherosclerosis, and Aortic Root Dilation
Recipients of SR-uPA+/0Plau+/+ BM began to die suddenly approximately 7 weeks after transplant (supplemental Figure III; P<0.001 for survival compared to recipients of SR-uPA0/0 Plau+/+ BM). Atherosclerosis was measured in recipients of the SR-uPA+/0Plau+/+ BM that survived for 14 weeks after transplantation. These mice had a 2-fold increase in percent Sudan IV-stained area on pinned aortae (P<0.001 versus recipients of SR-uPA0/0Plau+/+ BM; Figure 3A). Aortic root intimal lesion area was also increased 2.5-fold (P<0.001 versus recipients of SR-uPA0/0Plau+/+ BM; Figure 3B through 3D). In addition, mean luminal stenosis of the proximal coronary arteries was significantly more severe in recipients of SR-uPA+/0Plau+/+ BM (65±28 versus 34±19%; P=0.01; Table 1). Aortic root lesions of recipients of SR-uPA+/0Plau+/+ BM also had significantly more macrophage- and lipid-stained area. However, the percentage of total lesion area occupied by macrophages and lipids was equivalent in recipients of SR-uPA+/0Plau+/+ or SR-uPA0/0Plau+/+ BM (P≥0.2; Table 1).
Recipients of SR-uPA+/0Plau+/+ BM had dilated aortic roots and a small but significant increase in total aortic surface area (internal elastic lamina circumference increased by 20%; total aortic surface area by 5%; P≤0.006; Table 1). Aortic wall medial destruction was also more severe in recipients of SR-uPA+/0Plau+/+ BM. An observer blinded to BM donor genotype examined a single Movat-stained aortic root section of 12 recipient mice (6 from each group) and—based on the presence or absence of significant medial destruction—correctly identified the BM donor genotype for all 12 mice (P=0.002).
Transplantation of SR-uPA0/0Plau−/− Bone Marrow Reduces Late Aortic Atherosclerosis
We next tested whether loss of uPA expression in macrophages would decrease atherosclerosis. We first compared advanced atherosclerosis in 37-week-old fat-fed recipients of SR-uPA0/0Plau−/− and SR-uPA0/0Plau+/+ BM, all transplanted at 8 weeks of age. Recipients of SR-uPA0/0Plau−/− BM had significantly less Sudan IV–stained aortic surface area (both total and %) than recipients of SR-uPA0/0Plau+/+ BM (40% decrease; P=0.03; Figure 4A, Table 2). This decrease was more evident in the abdominal than the thoracic aorta (Figure 4B and 4C, supplemental Figure IVA and IVB). In contrast to the effect on aortic surface plaque area, transplantation of SR-uPA0/0Plau−/− BM did not decrease aortic root intimal area (P=0.3 for recipients of SR-uPA0/0 Plau−/− versus SR-uPA0/0Plau+/+ BM; Table 2, Figure 4D, supplemental Figure IVC and IVD). The 37-week-old SR-uPA0/0Plau−/− BM recipients also had unaltered aortic root circumference, aortic surface area, aortic root lesion lipid and macrophage area, and a similar severity of coronary artery stenoses (Table 2; P≥0.3 for all).
Finally, we tested whether macrophage-specific loss of uPA expression in recipients of SR-uPA0/0Plau−/− BM would also retard early lesion development. We compared atherosclerosis in chow-fed recipients of either SR-uPA0/0 Plau+/+ or SR-uPA0/0Plau−/− BM, transplanted at 8 weeks of age and harvested 7 weeks later. Atherosclerosis in these mice was limited, and there were no significant differences in aortic root intimal area, aortic surface area, Sudan IV–positive aortic surface area, aortic root circumference, aortic root macrophage area (total or percent), or aortic root lesion lipid area between recipients of SR-uPA0/0 Plau−/− and SR-uPA0/0Plau+/+ BM (Table 2, supplemental Figure IVE and IVF).
SR-uPA+/0 Macrophages Have Increased Migration but Not Increased Activation
To explore mechanisms for accelerated atherosclerosis in SR-uPA+/0 mice, we tested whether SR-uPA+/0 macrophages migrated more readily through extracellular matrix and whether SR-uPA+/0 macrophages were more responsive to activation by inflammatory stimuli. SR-uPA+/0 macrophages migrated through Matrigel more than nontransgenic macrophages (34 versus 11 migrated cells per high-power field; P=0.03; supplemental Figure V). However, SR-uPA+/0 and nontransgenic macrophages showed similar cytokine expression after treatment with LPS or oxidized LDL (supplemental Figure VI).
We used a BMT approach and donor mice with either macrophage-targeted uPA overexpression (SR-uPA+/0 mice) or genetic deficiency of uPA (Plau−/− mice) to investigate the role of macrophage-expressed uPA in atherosclerosis. Our major findings were: (1) BMT from SR-uPA+/0 donors yields mice with macrophage-specific uPA overexpression; (2) BMT from Plau−/− donors yields mice with macrophage-specific loss of uPA expression; (3) BMT from SR-uPA+/0 donors increases plasma uPA; however, BMT from Plau−/− donors does not decrease plasma uPA; (4) The uPA/plasminogen-dependent vascular and sudden death phenotypes of the SR-uPA+/0Apoe−/− mice are all adoptively transferred by BMT; (5) Postnatal deletion of uPA expression from BM-derived macrophages decreases aortic atherosclerosis. Our results suggest that the level of macrophage uPA expression—over a broad range—is an important determinant of atherosclerotic lesion growth.
Our initial characterization of the SR-uPA+/0 mice revealed high levels of SR-uPA transgene mRNA in peritoneal macrophages, and lower levels in heart, kidney, liver, lung, brain, aorta, and spleen.10 Because the human scavenger receptor promoter used to express uPA in the SR-uPA+/0 mice reportedly drives transgene expression predominantly in macrophages,23 we hypothesized that the presence of SR-uPA transgene mRNA in tissues such as heart, liver, and lung was attributable to blood-derived resident macrophages in these tissues. However, it was also possible that the scavenger receptor promoter was “leaky” and that SR-uPA mRNA in these tissues was within cell types other than macrophages. In support of this, uPA expression in some of our original SR-uPA founder lines was less specific to macrophage-rich tissues than the 3 lines we eventually studied.10 These founder lines were not maintained; however, they did raise concern that—depending on integration site—the SR-uPA transgene could be expressed in nonmacrophage cells. Expression of the SR-uPA transgene outside the artery wall, either by tissue macrophages or other cell types, raised the possibility that the vascular and sudden death phenotypes of SR-uPA+/0 mice might be caused by uPA expression at extra-arterial sites and not by uPA-overexpressing artery wall macrophages.
Here we used transplantation of SR-uPA+/0Plau+/+ BM into SR-uPA0/0Plau+/+ mice to test specifically whether macrophage uPA overexpression causes the vascular and sudden death phenotypes of SR-uPA+/0 mice. The essentially complete reproduction of the SR-uPA+/0Apoe−/− phenotypes in Apoe−/− recipients of SR-uPA+/0 BM (Figure 3 and supplemental Figure III, Table 1) largely excludes the possibility that these phenotypes are caused by uPA overexpression in nonhematopoietic cells. Moreover, our finding (supplemental Figure I and data not shown) that peripheral blood leukocytes do not express the SR-uPA transgene identifies uPA-overexpressing macrophages as the sole hematopoietic cell type that accelerates atherosclerosis and causes sudden death in SR-uPA+/0Apoe−/− mice. Although we cannot completely exclude a pathological role for elevated plasma uPA that unavoidably accompanies transplantation of SR-uPA+/0 BM (Figure 2), it seems more likely that the phenotypes of SR-uPA+/0 mice (ie, plaque growth, medial destruction, and sudden death) result from direct actions of uPA-overexpressing macrophages on the artery wall and are not caused by elevated plasma uPA. uPA-overexpressing macrophages are abundant in the artery wall and heart (sites at which SR-uPA+/0 mice have impressively altered phenotypes).10,22 In contrast elevated plasma uPA is present throughout the mouse, but appears to have few—if any—effects outside the heart and aorta.10,22 Nevertheless, the only means for definitively testing the hypothesis that elevated plasma uPA is sufficient to accelerate atherosclerosis would be to generate mice with elevated plasma uPA but without overexpression in macrophages.
It was heretofore uncertain that decreased macrophage uPA expression would be atheroprotective. Carmeliet et al found no difference in extent of aortic root atherosclerosis in Plau−/−Apoe−/− versus Plau+/+Apoe−/− mice, although lack of uPA was associated with preservation of the aortic root media.12 A recent study also reported unaltered aortic root atherosclerosis in Plau−/−Apoe−/− versus Plau+/+ Apoe−/− mice. However, Plau−/−Apoe−/− mice had marginally more severe coronary stenosis (90 versus 83%) and a trend toward increased brachiocephalic plaque area.13 Taken together, these studies portray uPA as an initiator of medial destruction but either without effect on atherosclerosis or marginally—and site-specifically—protective.22
In agreement with these precedents, postnatal loss of macrophage uPA expression does not affect aortic root atherosclerosis (Figure 4D, supplemental Figure IVC and IVD). However, loss of macrophage uPA expression substantially retards aortic atherosclerosis more distally (Figure 4A through 4C, supplemental Figure IVA and IVB). There are several possible reasons why we found an atheroprotective role for uPA deficiency and others did not. First, the previous studies did not examine atherosclerosis in pinned aortae and therefore may have missed this effect.12,13 Second, others used mice of a mixed genetic background12 and studied mice of different ages that were fed different diets (ie, 30 weeks/cholate-containing diet12 and 1 year/chow diet13 versus 37 weeks/Western diet used here). Third, in both previous studies the experimental Plau−/−Apoe−/− mice were completely deficient in uPA throughout life. In contrast, the Plau+/+Apoe−/− recipients of Plau−/−Apoe−/− BM in the present study lost uPA expression (in their macrophages only) during adult life. It is possible that lack of uPA expression during development is compensated by upregulation of other proteases or downregulation of inhibitors. Alternatively, in Plau−/−Apoe−/− mice effects attributable to loss of uPA expression in one cell type (eg, macrophages) could be countered by effects of loss of uPA expression in other cell types. Both compensatory changes during development and variability in cell-type-specific effects of uPA deficiency could obscure atheroprotective effects of loss of macrophage uPA expression in Plau−/−Apoe−/− mice. Variability between our study and its precedents might also be explained by the complete absence of plasma uPA in Plau−/− Apoe−/− mice, a deficiency associated with systemic consequences.14,24,25 In contrast, transplantation of Plau−/− BM in the present study does not lower plasma uPA of recipients (Figure 2) and is therefore much less likely to have systemic effects (this contrasts with the SR-uPA+/0 BMT results in which both macrophage uPA expression and plasma uPA levels are elevated in BM recipients). Preserved plasma uPA levels in Plau−/− BM recipients, combined with the lack of uPA expression in peripheral blood leukocytes of Plau+/+ mice, establishes that the only reproducible change in uPA expression after reconstitution of Plau+/+ mice with Plau−/− BM is loss of uPA expression by macrophages. Therefore, loss of uPA expression—by macrophages—reduces aortic atherosclerosis in recipients of Plau−/−Apoe−/− BM.
The absence of aortic root atheroprotection by transplanted Plau−/− BM might be explained by the presence of different, atypical mediators of atherosclerosis in this location compared to the remainder of the Apoe−/− vascular tree. Lack of an effect in the aortic root could also be explained by the timing of the BMT. Plau−/− BM engraftment at 8 to 12 weeks in fat-fed mice occurs after initiation of aortic root atherosclerosis and after accumulation—in the aortic root intima—of Plau+/+ macrophages.26 These atherogenic macrophages may be postmitotic and would therefore survive pre-BMT irradiation. In contrast, lesions in the distal aorta develop later, after BMT, and would be populated exclusively by less atherogenic donor-derived Plau−/− macrophages. Alternatively, the atheroprotective effects of Plau−/− BM might be attributable to a site-restricted role for uPA in atherosclerosis: uPA might accelerate atherosclerosis in the distal aorta but not in the aortic root. In this case, loss of macrophage-expressed uPA would not affect aortic root atherosclerosis. Site-restricted effects of molecular modifiers of atherosclerosis are well described but poorly understood.27 There is some experimental support for a site-restricted role for uPA: angiotensin II infusion causes uPA-dependent aneurysm formation in the abdominal aorta, but not in the aortic root.28
The atheroprotective effects of Plau−/− BM could also be stage-restricted. Transplantation of Plau−/− BM decreased atherosclerosis in 37-week-old fat-fed mice (with advanced lesions) but did not appear to retard atherosclerosis in younger, chow-fed recipients (with early lesions). Lack of an effect of Plau−/− BM on early lesions suggests a role for macrophage-derived uPA in lesion progression rather than initiation. A role for uPA in lesion progression is consistent with 2 other observations: (1) monocytes (initiators of plaques) do not express uPA whereas macrophages (residents of plaques) do; and (2) in SR-uPA+/0 mice, uPA overexpression has a more robust effect on longstanding aortic root lesions than on later-developing more distal aortic lesions.10,11
Our data begin to address mechanisms through which macrophage uPA expression accelerates atherosclerosis. Absence of uPA-related differences in percentages of lesions occupied by macrophages argues against a role for uPA in monocyte/macrophage homing. However, small differences might not be detected, and more sensitive assays for monocyte homing would address this more definitively.29 We also found no evidence that uPA-overexpressing macrophages are more easily activated; however, uPA-overexpressing macrophages migrate more readily through matrix (supplemental Figure V), which could contribute to lesion growth. Increased in vitro macrophage migration might not be accompanied by an increased percentage of lesion macrophages because the in vivo assay might not be sufficiently sensitive to detect small differences in lesion composition. Alternatively, accumulation of other lesion components (lipid, matrix) might also be accelerated in uPA-overexpressing mice.
In summary, the level of uPA expression by macrophages is rate-limiting for atherosclerotic lesion growth in Apoe−/− mice. Macrophage uPA expression and associated downstream plasminogen-dependent11 processes (potentially including aneurysm formation as well as plaque rupture)12,30 may be useful targets for a single therapy, targeted at uPA, that would both slow the progression of human atherosclerosis and prevent its complications.
We thank Alyssa Wu-Zhang and Amelia Buben for technical assistance and Margo Weiss for assistance with manuscript preparation.
Sources of Funding
This work was supported by National Heart, Lung, and Blood Institute Grants R01 HL080597 (to D.A.D.) and T32 HL07828 (to R.K. and M.K.). S.D.F. was supported by a Howard Hughes Medical Student Research Fellowship. K.I.S. was supported in part by a grant to the University of Washington from the Howard Hughes Medical Institute through the Undergraduate Science Education Program. D.A.D. was also supported by the John L. Locke Jr Charitable Trust.
R.K. and M.K. contributed equally to this study.
Received January 30, 2009; revision accepted August 20, 2009.
Raghunath PN, Tomaszewski JE, Brady ST, Caron RJ, Okada SS, Barnathan ES. Plasminogen activator system in human coronary atherosclerosis. Arterioscler Thromb Vasc Biol. 1995; 15: 1432–1443.
Folsom AR, Aleksic N, Park E, Salomaa V, Juneja H, Wu KK. Prospective study of fibrinolytic factors and incident coronary heart disease: the Atherosclerosis Risk in Communities (ARIC) Study. Arterioscler Thromb Vasc Biol. 2001; 21: 611–617.
Cushman M, Lemaitre RN, Kuller LH, Psaty BM, Macy EM, Sharrett AR, Tracy RP. Fibrinolytic activation markers predict myocardial infarction in the elderly. The Cardiovascular Health Study. Arterioscler Thromb Vasc Biol. 1999; 19: 493–498.
Sakkinen PA, Cushman M, Psaty BM, Rodriguez B, Boineau R, Kuller LH, Tracy RP. Relationship of plasmin generation to cardiovascular disease risk factors in elderly men and women. Arterioscler Thromb Vasc Biol. 1999; 19: 499–504.
Cozen AE, Moriwaki H, Kremen M, DeYoung MB, Dichek HL, Slezicki KI, Young SG, Veniant M, Dichek DA. Macrophage-targeted overexpression of urokinase causes accelerated atherosclerosis, coronary artery occlusions, and premature death. Circulation. 2004; 109: 2129–2135.
Kremen M, Krishnan R, Emery I, Hu JH, Slezicki KI, Wu A, Qian K, Du L, Plawman A, Stempien-Otero A, Dichek DA. Plasminogen mediates the atherogenic effects of macrophage-expressed urokinase and accelerates atherosclerosis in apoE-knockout mice. Proc Natl Acad Sci U S A. 2008; 105: 17109–17114.
Fazio S, Babaev VR, Murray AB, Hasty AH, Carter KJ, Gleaves LA, Atkinson JB, Linton MF. Increased atherosclerosis in mice reconstituted with apolipoprotein E null macrophages. Proc Natl Acad Sci U S A. 1997; 94: 4647–4652.
Otsuka G, Agah R, Frutkin AD, Wight TN, Dichek DA. Transforming growth factor beta 1 induces neointima formation through plasminogen activator inhibitor-1-dependent pathways. Arterioscler Thromb Vasc Biol. 2006; 26: 737–743.
Fazio S, Yao Z, McCarthy BJ, Rall SC Jr. Synthesis and secretion of apolipoprotein E occur independently of synthesis and secretion of apolipoprotein B-containing lipoproteins in HepG2 cells. J Biol Chem. 1992; 267: 6941–6945.
Pitas RE, Boyles JK, Lee SH, Hui D, Weisgraber KH. Lipoproteins and their receptors in the central nervous system. Characterization of the lipoproteins in cerebrospinal fluid and identification of apolipoprotein B,E(LDL) receptors in the brain. J Biol Chem. 1987; 262: 14352–14360.
Groeneweg M, Kanters E, Vergouwe MN, Duerink H, Kraal G, Hofker MH, de Winther MP. Lipopolysaccharide-induced gene expression in murine macrophages is enhanced by prior exposure to oxLDL. J Lipid Res. 2006; 47: 2259–2267.
Moriwaki H, Stempien-Otero A, Kremen M, Cozen AE, Dichek DA. Overexpression of urokinase by macrophages or deficiency of plasminogen activator inhibitor type 1 causes cardiac fibrosis in mice. Circ Res. 2004; 95: 637–644.
Horvai A, Palinski W, Wu H, Moulton KS, Kalla K, Glass CK. Scavenger receptor A gene regulatory elements target gene expression to macrophages and to foam cells of atherosclerotic lesions. Proc Natl Acad Sci U S A. 1995; 92: 5391–5395.
Levi M, Moons L, Bouche A, Shapiro SD, Collen D, Carmeliet P. Deficiency of urokinase-type plasminogen activator-mediated plasmin generation impairs vascular remodeling during hypoxia-induced pulmonary hypertension in mice. Circulation. 2001; 103: 2014–2020.
Lluis F, Roma J, Suelves M, Parra M, Aniorte G, Gallardo E, Illa I, Rodriguez L, Hughes SM, Carmeliet P, Roig M, Munoz-Canoves P. Urokinase-dependent plasminogen activation is required for efficient skeletal muscle regeneration in vivo. Blood. 2001; 97: 1703–1711.
Nakashima Y, Plump AS, Raines EW, Breslow JL, Ross R. ApoE-deficient mice develop lesions of all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994; 14: 133–140.
VanderLaan PA, Reardon CA, Getz GS. Site specificity of atherosclerosis: site-selective responses to atherosclerotic modulators. Arterioscler Thromb Vasc Biol. 2004; 24: 12–22.
Deng GG, Martin-McNulty B, Sukovich DA, Freay A, Halks-Miller M, Thinnes T, Loskutoff DJ, Carmeliet P, Dole WP, Wang YX. Urokinase-type plasminogen activator plays a critical role in angiotensin II-induced abdominal aortic aneurysm. Circ Res. 2003; 92: 510–517.
Tacke F, Ginhoux F, Jakubzick C, van Rooijen N, Merad M, Randolph GJ. Immature monocytes acquire antigens from other cells in the bone marrow and present them to T cells after maturing in the periphery. J Exp Med. 2006; 203: 583–597.
Slager CJ, Wentzel JJ, Gijsen FJ, Thury A, van der Wal AC, Schaar JA, Serruys PW. The role of shear stress in the destabilization of vulnerable plaques and related therapeutic implications. Nat Clin Pract Cardiovasc Med. 2005: 456–464.