Uremia-Specific Effects in the Arterial Media During Development of Uremic Atherosclerosis in Apolipoprotein E-Deficient Mice

【摘要】  Objective- Uremia accelerates formation of atherosclerosis-like lesions in apolipoprotein E-deficient (apoE -/- ) mice. In this study, we compared gene expression patterns in classical and uremic atherosclerosis.

Methods and Results- High-density oligonucleotide microarray analyses were performed with aortic RNA from 5/6 nephrectomized (NX) and sham-operated mice. After 12 weeks, NX apoE -/- mice had more atherosclerosis and 24 genes were differentially expressed as compared with sham apoE -/- mice. Nine genes expressed in muscle cells displayed reduced expression (3.3- to 142-fold, P <0.05), whereas osteopontin gene expression was increased 8.7-fold ( P <0.05) in NX mice. Studies of NX wild-type mice suggested that the changes in NX apoE -/- mice were dependent on hypercholesterolemia. Nevertheless, lesioned versus nonlesioned areas of aortas from nonuremic apoE -/- mice with classical atherosclerosis displayed less pronounced reductions in expression of the muscle cell related genes than seen in NX apoE -/- mice even though the osteopontin gene expression was increased &15-fold. Electron microscopy showed more vacuolized and necrotic smooth muscle cells within the media underneath both nonlesioned and lesioned intima in NX than in sham apoE -/- mice.

Conclusion- The results suggest that uremic vasculopathy in apoE -/- mice, in addition to intimal atherosclerosis, is characterized by a uremia-specific medial smooth muscle cell degeneration, which appears to be accentuated by hypercholesterolemia.

To compare the molecular pathophysiology of classical and uremic atherosclerosis, we used high-density oligonucleotide microarray analysis to assess gene expression patterns in aortas from 5/6 nephrectomized and sham-operated apolipoprotein E-deficient mice. The results reveal that uremic vasculopathy, in addition to intimal atherosclerosis, is characterized by specific medial smooth muscle cell degeneration.

【关键词】  renal failure uremia atherosclerosis vascular smooth muscle cells apolipoprotein Edeficient mice

Introduction

The cardiovascular mortality rate is increased &300-fold in 30-year-old individuals on dialysis compared with individuals without renal disease, 1 and even mild renal failure is an independent risk factor of cardiovascular disease. 2,3 Although renal dysfunction often is accompanied by dyslipidemia, hypertension, and diabetes, the high prevalence of cardiovascular disease in patients with renal disease cannot be explained by the classical risk factors alone. 4,5 Thus, renal dysfunction is also associated with nontraditional cardiovascular risk factors, including activation of the renin angiotensin system. 6

Even though renal failure is accompanied by calcification of arteries, 7 it has been controversial whether uremia might accelerate formation of atherosclerosis. Recent experiments from three different groups (including ours) 8-10 have disclosed that 5/6 nephrectomy (NX) in apolipoprotein E-deficient (apoE -/- ) mice accelerates formation of intimal lesions in aorta. NX apoE -/- mice develop 6- to 10-times more lesions than sham-operated control mice despite similar blood pressure and plasma homocysteine concentrations, and only minor differences in plasma lipoprotein concentrations. 8,11 The NX mice are also characterized by a low body weight, reduced blood hemoglobin, and increased plasma calcium x phosphate product. 8,11

Intimal lesions in NX mice are qualitatively similar to those in classical atherosclerosis, ie, with accumulation of macrophage-derived foam cells 8-10 and cholesteryl esters. 11 In addition, there is pronounced accumulation of nitrotyrosine (a marker of reactive oxygen species-protein interactions) in the uremic lesions, 8,9 and lesion formation in NX apoE -/- mice is preceded by upregulation of intercellular adhesion molecule-1 (ICAM-1) expression and accompanied by upregulation of vascular cell adhesion molecule (VCAM)-1 expression in the arterial wall 11. Nevertheless, the culprit of uremic atherosclerosis remains to be identified.

In this study we sought to obtain further knowledge on the molecular pathophysiology of vascular changes caused by NX in apoE -/- mice by performing gene expression profiling studies of aortas from NX and sham mice. The results revealed that uremic compared with classical atherosclerosis is accompanied by decreased expression of muscle cell genes and accelerated medial smooth muscle cell degeneration in apoE -/- mice.

Materials and Methods

Please see http://atvb.ahajournals.org.

Results

Aortic Gene Expression 12 Weeks After 5/6 Nephrectomy

Subtotal 5/6 nephrectomy (NX) of apoE -/- mice caused a 103% increase of plasma urea (23.7±0.6 versus 11.7±0.3 mmol/L; P <0.001) and a 57% increase of plasma creatinine (47±1 versus 30±1 µmol/L; P <0.001). 11 The plasma calcium x phosphate product was 38% higher (7.38±0.58 versus 5.35±0.18 mmol/L 2; P <0.01), 11 whereas the plasma cholesterol concentration increased only slightly after NX (13.1±0.5 versus 11.3±0.4 mmol/L; P <0.05). In a parallel study, the total aortic plaque area fraction as determined by digital image analysis of the intimal surface of the aorta from the heart to the iliac arteries was larger in NX (3.9±0.5%) than in sham apoE -/- mice (0.4±0.1%, P <0.0001) when examined 12 weeks after surgery. 11

To examine the gene expression profiles in aortas, we hybridized 3 pools of aortic cRNA (each made with RNA from 3 mouse aortas) from NX mice and 3 pools from sham mice to high-density oligonucleotide 1.5-fold and an absolute change in expression of 50 U with P <0.05) and 12 genes showed decreased expression in NX compared with sham mice (Table I, available online at http://atvb.ahajournals.org). The biological functions of the differentially expressed genes were predominantly assigned to muscle structure and development, inflammation, extracellular matrix remodeling, or metabolism. The most prominent changes were 3.3- to 142-fold downregulations of transcripts assigned to muscle structure and development, eg, parvalbumin, myosin, and -actin encoding genes. Transcripts involved in inflammation were all upregulated (1.8- to 8.7-fold) and included osteopontin, matrix metalloproteinase(MMP)-3 and -12, VCAM-1, and serum amyloid A. The changes were verified by real-time polymerase chain reaction (PCR) quantification of 8 selected genes ( Figure 1 ).

Figure 1. High-density oligonucleotide microarray and real-time PCR analysis of aortic gene expression 12 weeks after 5/6 nephrectomy in apoE -/- mice. Fold-changes in aortic expression of 8 genes in 5/6 nephrectomized (NX) versus sham-operated apoE -/- mice were determined by high density oligonucleotide microarray analysis (black bars) with 3 pools of RNA (each with RNA from 3 mice) from each group or real-time PCR (white bars) with RNA from each mouse aorta (n=9 x 2). SPLI, secretory leukocyte protease inhibitor.

The effect of NX on gene expression in the aorta was also assessed in normocholesterolemic wild-type mice. Twelve weeks after NX or sham operation, the plasma urea concentration was 22.7±1.5 versus 10.6±0.8 mmol/L in NX as compared with sham wild-type mice ( P <0.0001), whereas the plasma calcium x phosphate product (5.05±0.13 versus 5.08±0.19 mmol/L 2 ) and the cholesterol concentration (2.4±0.1 versus 2.3±0.1 mmol/L) did not differ between the 2 groups. No arterial lesions were seen in the aortas of these mice. There were no significant differences in the expression levels of transcripts assigned to inflammation (osteopontin and MMP-12) or muscle cell biology (myosin, -actin, and troponin) in NX compared with sham aortas from the wild-type mice ( Figure 2 ).

Figure 2. Patterns of aortic gene expression in uremic and classical atherosclerosis. Fold-changes in aortic gene expression were determined with real-time PCR in 5/6 nephrectomized (NX) versus sham-operated apoE -/- mice at 12 weeks (white bars, n=9 x 2) or at 2 weeks (black bars, n = 9 x 2) after surgery, in NX versus sham wild-type mice at 12 weeks after surgery (hatched bars, n=18 and n=9, respectively), and in lesioned versus nonlesioned areas of aortas from apoE -/- mice with classical atherosclerosis (gray bars, n = 4 x 2). All indicated genes displayed statistically significant changes in expression ( P <0.05) in NX versus sham apoE -/- mice after 12 weeks, whereas the expression levels of all genes were similar in NX versus sham apoE -/- mice after 2 weeks and in NX versus sham wild-type mice after 12 weeks. In lesioned versus nonlesioned areas of aortas from mice with classical atherosclerosis the expression levels of osteopontin, MMP-12, and glycoprotein 49A genes were significantly higher in the lesioned areas, whereas the changes in SLPI, MMP-3, actin, myosin, and troponin gene expression were not statistically significant. SPLI, secretory leukocyte protease inhibitor.

Aortic Gene Expression in Classical Atherosclerosis

To assess to what extent the observed changes in aortic gene expression in NX apoE -/- mice reflected more advanced atherosclerosis than in sham apoE -/- mice or uremia-specific effects, we examined RNA from lesioned and nonlesioned areas of aortas from nonuremic apoE -/- mice with classical atherosclerosis. The plaque area fraction was &30% in the lesioned areas versus &0% in the nonlesioned areas; 12 both lesioned and nonlesioned segments were isolated from the aortic arch, the thoracic aorta, and the abdominal aorta. On real-time PCR, the expression of osteopontin, MMP-12, and glycoprotein 49A mRNAs were markedly increased in lesioned compared with nonlesioned areas. The increases were more pronounced in lesioned versus nonlesioned areas of aortas with classical atherosclerosis than in NX versus sham apoE -/- aortas 12 weeks after surgery (15-fold versus 8.7-fold for osteopontin, 28-fold versus 2.8-fold for MMP-12, and 62-fold versus 2.3-fold for glycoprotein 49A) ( Figure 2 ). Nevertheless, the expression of the mRNAs for -actin, myosin, and troponin was reduced to a similar or lesser extent (and not statistically significantly) in lesioned versus nonlesioned areas of aortas from apoE -/- mice with classical atherosclerosis than in NX versus sham apoE -/- mouse aortas.

Aortic Gene Expression 2 Weeks After 5/6 Nephrectomy

We assessed changes in gene expression that preceded the development of atherosclerosis in NX apoE -/- mice. Two weeks after surgery the NX apoE -/- aortas showed no lesions. 11 There was no difference in the expression of the 9 transcripts assigned to muscle cell biology that were downregulated after 12 weeks of uremia ( Figure 2 ). After 2 weeks, 37 genes showed 1.5-fold) and 4 genes showed decreased expression in NX compared with sham apoE -/- mice (Table II, available online at http://atvb.ahajournals.org). The differentially expressed genes were assigned to inflammation, signal transduction, cell differentiation and growth, extracellular matrix remodeling, metabolism, or transporter activity. Two of the genes involved in inflammation (MMP-3 and glycoprotein 49A) were upregulated both after 2 and 12 weeks (Table II; Figure 2 ). Although 16 of the genes that showed increased expression after 2 weeks are involved in immunoglobulin production (Table II), we did not see IgG or B-lymphocyte accumulation in the arterial wall on immunostaining of nonlesioned NX apoE -/- aortas (data not shown). (The antibody used for staining of B-cells in aortic sections (CD22,2, catalog no. 553382; Pharmingen, San Diego, Calif) worked well on the positive controls (mouse spleen tissue). 8 However, because the number of B-cells in mouse aortas is usually very low, the negative finding may simply be a problem of detection.)

Expression of Osteopontin and -Actin Protein in 5/6 Nephrectomized Mouse Aortas

Osteopontin constituted the most markedly upregulated transcript in NX versus sham apoE -/- mouse aortas 12 weeks after surgery (Table I). In 12-week NX apoE -/- aortas, osteopontin protein staining was predominantly associated with macrophages in intimal lesions ( Figure 3 A). Osteopontin protein was also detected in and around the medial smooth muscle cells underneath the intimal lesions.

Figure 3. Immunohistochemical visualization of aortic -actin and osteopontin in 5/6 nephrectomized mice. Sections are from the aortic root of NX mice 12 weeks after nephrectomy. A, Immunohistochemistry showing osteopontin staining associated with intimal macrophages and medial smooth muscle cells beneath intimal lesions (L). B, Immunohistochemistry showing loss of -actin staining in the media under intimal lesions. C, Elastin (Verhoeff) staining showing destruction and disorganization of elastic laminae beneath intimal lesions. D, Immunohistochemistry without a primary antibody (negative control). Original magnification, 200 x (A) and 25 x (B, C, and D).

Twelve-week NX apoE -/- mouse aortas displayed loss of -actin protein staining underneath intimal lesions ( Figure 3 B), whereas the medial staining pattern in nonlesioned parts of the aortic wall could not be distinguished from that in sham apoE -/- mice (data not shown). Elastin staining showed disruption and loss of organization of the elastic laminae underneath the intimal lesions as compared with the media underneath nonlesioned intima in NX and sham apoE -/- mice ( Figure 3 C).

Ultrastructure of Aortic Media in 5/6 Nephrectomized Mice With Advanced Atherosclerosis

To further evaluate the structural characteristics of the media in NX apoE -/- mice, we examined proximal aortas from 3 NX and 3 sham apoE -/- mice with advanced atherosclerosis (37 weeks after NX) with electron microscopy ( Figure 4 ). The intimal lesions in NX and sham apoE -/- aortas both contained cholesterol crystals and lipid-filled macrophages ( Figure 4A and 4 B). However, the ultrastructure of the media was different in the NX compared with the sham apoE -/- mice, both underneath nonlesioned and lesioned intima. In the nonlesioned areas, the media of NX apoE -/- aortas was characterized by fewer smooth muscle cells and more intercellular matrix than seen in sham apoE -/- aortas. Also, in NX apoE -/- mice more of the smooth muscle cells appeared irregularly shaped and/or vacuolized in comparison with the smooth muscle cells in media under nonlesioned intima in the apoE -/- sham mice ( Figure 4E and 4 C). The media underneath intimal lesions of NX apoE -/- mice was characterized by extensive smooth muscle cell vacuolization and scattered necrosis ( Figure 4F and 4 G). Many of the smooth muscle cells in NX apoE -/- mice appeared slender with large projections ( Figure 4 H). Accordingly, the amount of intercellular substance appeared markedly increased. These changes in media underneath intimal lesions were all more pronounced in NX than in sham apoE -/- mice. Of note, we did not see calcifications in the media or intimal lesions on inspections 80 aortic sections with the electron microscope.

Figure 4. Electron microscopy showing altered ultrastructure of aortic media in 5/6 nephrectomized mice. Ultrathin sections of the aortic root from 5/6 nephrectomized (NX) and sham-operated apoE -/- mice were examined 37 weeks after surgery. A and B, Intimal lesion in a NX mouse aorta with cholesterol crystals (A) and lipid-filled macrophages (B). C, media under nonlesioned intima (i) in a sham mouse aorta. D, media under lesioned intima in a sham mouse aorta. Note, the more vacuolized and irregularly shaped smooth muscle cells and increased deposition of intercellular matrix as compared with C (media under nonlesioned intima in sham mouse). E, Media under nonlesioned intima (i) in a NX mouse aorta. The density of the smooth muscle cells was decreased with more intercellular matrix and some of the smooth muscle cells were irregularly shaped and vacuolized, as compared with (C) (media under nonlesioned intima in sham mouse). F to H, Media under lesioned intima in NX mouse aorta. In comparison with media under nonlesioned intima in NX mice (E), as well as media under intimal lesions in sham mice (D), the smooth muscle cells were more vacuolized and surrounded by increased intercellular matrix (F). Under lesioned intima in NX mouse aorta some smooth muscle cells were necrotic (G), and others were irregularly shaped with large projections; those alterations were not seen under nonlesioned intima in NX mouse aortas or in sham mouse aortas. Original magnification 3500 x (A, D, F, G), 2200 x (E), 2800 x (C and H), and 4400 x (B).

Discussion

In this study, we compared gene and protein expression and ultrastructural characteristics of the aortas in NX and sham apoE -/- mice to gain new insight into putative uremia-specific effects on the arterial wall during development of atherosclerosis.

Despite a &10-fold larger atherosclerotic lesion area in NX compared with sham apoE -/- mice 12 weeks after surgery, 11 1.5-fold changes in expression level between NX and sham apoE -/- mouse aortas. A priori, it could be suspected that some of the observed changes simply would result from multiple comparisons. However, several of the genes displaying differential expression are implicated in atherosclerosis (eg, osteopontin, MMP-3, MMP-12, VCAM-1, serum amyloid A). 11,13-15 Also, for 8 selected genes the changes in transcript expression were verified by real-time PCR analysis. Because the NX apoE -/- mice contained more foam cell lesions than sham apoE -/- mice after 12 weeks, it was not surprising to see an upregulation of genes that are expressed by macrophages in atherosclerotic lesions. However, it was surprising that 9 of 24 differentially expressed genes could be assigned to muscle cell function and that all displayed markedly decreased expression.

In contrast to the finding in the hypercholesterolemic apoE -/- mice, the aortic expression of smooth muscle cell genes did not differ between NX and sham normocholesterolemic mice. This finding may reflect that hypercholesterolemia is required to change smooth muscle cell gene expression in uremic mouse aortas and raises the possibility that the changes in smooth muscle cell biology may be associated with the accelerated intimal lesion formation.

Classical atherosclerosis in nonuremic apoE -/- mice is accompanied by expansive remodeling of the media with disruption of elastic laminae and thinning of the smooth muscle cell layer. 16 The medial changes appear to become progressively pronounced with increasing growth of the intimal lesion 16 and have been associated with MMP actions. 17,18 However, the 3.3-fold downregulation of expression of smooth muscle cell assigned genes (corresponding to 70% reductions) suggests that NX in apoE -/- mice affects smooth muscle cells at all sites of the arterial wall and not only beneath the lesions because only 3.9% of the aortic surface area contained lesions. 11

To expand this idea, we compared the expression of both macrophage assigned and muscle cell assigned genes in lesioned and nonlesioned areas of the aortas in apoE -/- mice with classical atherosclerosis. Of note, in the lesioned areas &30% was lesion-covered, and the lesions were more advanced than those in the 12-week NX apoE -/- mice. 11 Accordingly, macrophage assigned genes displayed more pronounced increases in expression in lesioned versus nonlesioned areas of the aortas from apoE -/- mice with classical atherosclerosis than in the NX versus sham apoE -/- aortas after 12 weeks. In contrast, the downregulation of muscle cell assigned genes was similar or more pronounced in the NX versus sham apoE -/- aortas as compared with lesioned versus nonlesioned areas of nonuremic apoE -/- mouse aortas. These observations imply a uremia-specific affection of the arterial media that exceeds the affection of the media that occurs in association with classical atherosclerotic lesions. We cannot, however, conclude whether the changes in gene expression at 12 weeks cause subsequent loss of smooth muscle cells or reflect early smooth muscle cell affection by uremia that precedes the extensive damage seen at 37 weeks after NX.

Immunostaining of one of the downregulated gene products, ie, -actin, revealed a consistent loss of -actin beneath the early intimal lesions of NX apoE -/- mice. Also, elastin staining revealed disruption of elastic membranes underneath intimal lesions in NX apoE -/- mice. In this respect the development of intimal lesions is accompanied by changes in the media that are similar to those seen in classical atherosclerosis. 16 Electron microscopy of the media underneath lesioned intima showed alterations of the smooth muscle cell morphology both in NX and sham apoE -/- mice. Although the severely distorted smooth muscle cells in NX aortas could reflect advanced/extensive lesions, the electron microscopy of aortic media underneath nonlesioned intima in NX apoE -/- mice also revealed fewer smooth muscle cells with more of the cells characterized by irregular shaping and vacuolization compared with those in sham apoE -/- mice. The latter observations at 37 weeks agree with the results of the gene expression analysis at 12 weeks and support the notion that uremia affects the smooth muscle cells in the aortic media. Also, the morphological alterations in the media are similar to those described by Ejerblad et al 19 in radial arteries from uremic patients. It is unknown whether the medial affection may contribute to formation of atherosclerotic lesions in NX apoE -/- mice or whether they reflect an independent effect. A multitude of interventions that affect the arterial smooth muscle cell layer, eg, mechanical injury, 20 pharmacological manipulation of the renin-angiotensin system, 21 and genetic manipulation of smooth muscle cell genes 22 interferes with formation of intimal atherosclerotic lesions. This illustrates that alterations in the biology of the media can precipitate the formation of intimal lesions. Ejerblad et al 23,24 observed changes in the aortic media of NX rats without concomitant intimal lesions. Nevertheless, we did not observe changes in muscle cell biology related genes in NX wild-type mice. This may reflect that hypercholesterolemia is an important cofactor for the medial affection in NX ApoE -/- mice. Indeed, NX confers increased plasma cholesterol in rats. 24 It is also possible that other metabolic disturbances, eg, abnormalities in calcium and phosphate metabolism, may play a role. NX apoE -/- mice 11 and NX rats 24 display increased plasma calcium x phosphate products, whereas the calcium x phosphate product was not increased by NX in the C57Bl/6J wild-type mice in the present study.

In humans, uremia is accompanied by vascular calcifications and calcification of atherosclerotic lesions occurs in apoE -/- mice between 45 and 75 weeks of age 25. Interestingly, although the male NX apoE -/- mouse model has increased plasma calcium x phosphate product, 11 we did not see signs of accelerated calcification in the medial or intimal lesions. This is in accordance with previous studies from our group 8,11 and Buzello et al. 9 Nevertheless, a recent study by Massy et al 10 showed accentuated calcium deposition both in the intima and media of apoE -/- mice as early as 8 weeks after NX. A putative explanation may be a higher vitamin D 3 content and calcium to phosphate ratio in the mouse diet used by the latter group, or perhaps the inclusion of female mice in their study, since Massy et al 10 found that female mice had faster progression of vascular calcification than their male counterparts. Radial artery specimen from uremic patients display medial calcium deposits in and around extracellular vesicles. 19 Although we could see similar vesicular structures in the media on electron microscopy of aortas from apoE -/- mice 37 weeks after NX, we did not see any calcium deposition. The cause of the lack of calcifications in the male NX apoE -/- mice is not clear. It may be related to the marked upregulation of osteopontin which was seen both within the intimal lesions and in the media underneath the lesions. Osteopontin is believed to inhibit mineralization in bone 26 and vascular tissue. 27 Accordingly, excessive vascular calcifications develop in osteopontin-deficient male apoE -/- mice. 28

Gene expression analysis of aortas 2 weeks after NX in apoE -/- mice revealed that the changes in expression of muscle cell assigned genes had not occurred after 2 weeks. The plasma urea concentration is elevated and remains stable from 2 to 22 weeks after NX. 8 Thus, the media affection probably does not result from acute gene regulation by putative uremic toxins or inflammatory mediators, but rather is a progressive chronic process, perhaps reflecting continuous deposition of uremic metabolites in the arterial media, eg, advanced glycation end products or oxidized low density lipoprotein. Nevertheless, after 2 weeks the expression of several genes related to inflammation was upregulated in the NX apoE -/- aortas. This observation is in close accordance with a recent finding that upregulation of ICAM-1 precedes the development of uremic atherosclerosis. 11 Of note, after 2 weeks ICAM-1 was increased 1.5-fold in the microarray analysis, which is in agreement with our previous report. 11 It was striking that 16 of the 41 differentially expressed genes after 2 weeks were related to production of immunoglobulins (all were increased). Although we did not see any B-lymphocytes in sections of nonlesioned NX aortas, the result could reflect that uremia is accompanied by a humoral immune response. If this hypothesis is true, it will be of interest to assess the immune response against putative antigens induced by NX, eg, proteins modified by advanced glycation end products and oxidized low density lipoprotein. Both modifications may be immunogenic and multiple studies have shown that autoantibody formation against oxidized low density lipoprotein affects development of atherosclerosis. 29

In conclusion, the present study have provided novel insight into the molecular pathology of uremic atherosclerosis and implies that uremia in apoE -/- mice, in addition to markedly accelerating the formation of intimal lesions resembling classical atherosclerosis, also has pronounced specific effects on the arterial media smooth muscle cells.

Acknowledgments

The Danish Medical Research Council, The Danish Heart Foundation, The Copenhagen Hospital Corporation Research Council, The Danish Kidney Foundation, and The Helen and Ejnar Bjoernow Foundation supported this study. We thank Anne Andersen, Kirsten Bang, Annemette Borch, Nina Broholm, Karen Rasmussen, and Susanne Smed for technical assistance at various stages of this project.

【参考文献】
  Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis. 1998; 32: S112-S119.

Weiner DE, Tighiouart H, Amin MG, Stark PC, MacLeod B, Griffith JL, Salem DN, Levey AS, Sarnak MJ. Chronic kidney disease as a risk factor for cardiovascular disease and all-cause mortality: a pooled analysis of community-based studies. J Am Soc Nephrol. 2004; 15: 1307-1315.

Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, Parfrey P, Pfeffer M, Raij L, Spinosa DJ, Wilson PW. Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention. Circulation. 2003; 108: 2154-2169.

Cheung AK, Sarnak MJ, Yan G, Dwyer JT, Heyka RJ, Rocco MV, Teehan BP, Levey AS. Atherosclerotic cardiovascular disease risks in chronic hemodialysis patients. Kidney Int. 2000; 58: 353-362.

Longenecker JC, Coresh J, Powe NR, Levey AS, Fink NE, Martin A, Klag MJ. Traditional cardiovascular disease risk factors in dialysis patients compared with the general population: the CHOICE Study. J Am Soc Nephrol. 2002; 13: 1918-1927.

Zoccali C, Mallamaci F, Tripepi G. Novel cardiovascular risk factors in end-stage renal disease. J Am Soc Nephrol. 2004; 15 Suppl 1: S77-S80.

Goldsmith D, Ritz E, Covic A. Vascular calcification: a stiff challenge for the nephrologist: does preventing bone disease cause arterial disease? Kidney Int. 2004; 66: 1315-1333.

Bro S, Bentzon JF, Falk E, Andersen CB, Olgaard K, Nielsen LB. Chronic renal failure accelerates atherogenesis in apolipoprotein E-deficient mice. J Am Soc Nephrol. 2003; 14: 2466-2474.

Buzello M, Tornig J, Faulhaber J, Ehmke H, Ritz E, Amann K. The apolipoprotein E knockout mouse: a model documenting accelerated atherogenesis in uremia. J Am Soc Nephrol. 2003; 14: 311-316.

Massy ZA, Ivanovski O, Nguyen-Khoa T, Angulo J, Szumilak D, Mothu N, Phan O, Daudon M, Lacour B, Drueke TB, Muntzel MS. Uremia accelerates both atherosclerosis and arterial calcification in apolipoprotein E knockout mice. J Am Soc Nephrol. 2005; 16: 109-116.

Bro S, Moeller F, Andersen CB, Olgaard K, Nielsen LB. Increased expression of adhesion molecules in uremic atherosclerosis in apolipoprotein-E-deficient mice. J Am Soc Nephrol. 2004; 15: 1495-1503.

Moeller F, Nielsen LB. Aortic recruitment of blood lymphocytes is most pronounced in early stages of lesion formation in apolipoprotein-E-deficient mice. Atherosclerosis. 2003; 168: 49-56.

Lewis KE, Kirk EA, McDonald TO, Wang S, Wight TN, O?Brien KD, Chait A. Increase in serum amyloid a evoked by dietary cholesterol is associated with increased atherosclerosis in mice. Circulation. 2004; 110: 540-545.

Bruemmer D, Collins AR, Noh G, Wang W, Territo M, Arias-Magallona S, Fishbein MC, Blaschke F, Kintscher U, Graf K, Law RE, Hsueh WA. Angiotensin II-accelerated atherosclerosis and aneurysm formation is attenuated in osteopontin-deficient mice. J Clin Invest. 2003; 112: 1318-1331.

Jeng AY, Chou M, Sawyer WK, Caplan SL, Linden-Reed J, Jeune M, Prescott MF. Enhanced expression of matrix metalloproteinase-3, -12, and -13 mRNAs in the aortas of apolipoprotein E-deficient mice with advanced atherosclerosis. Ann N Y Acad Sci. 1999; 878: 555-558.

Bentzon JF, Pasterkamp G, Falk E. Expansive remodeling is a response of the plaque-related vessel wall in aortic roots of apoE-deficient mice: an experiment of nature. Arterioscler Thromb Vasc Biol. 2003; 23: 257-262.

Pasterkamp G, Schoneveld AH, Hijnen DJ, de Kleijn DP, Teepen H, van der Wal AC, Borst C. Atherosclerotic arterial remodeling and the localization of macrophages and matrix metalloproteases 1, 2 and 9 in the human coronary artery. Atherosclerosis. 2000; 150: 245-253.

Schoenhagen P, Vince DG, Ziada KM, Kapadia SR, Lauer MA, Crowe TD, Nissen SE, Tuzcu EM. Relation of matrix-metalloproteinase 3 found in coronary lesion samples retrieved by directional coronary atherectomy to intravascular ultrasound observations on coronary remodeling. Am J Cardiol. 2002; 89: 1354-1359.

Ejerblad S, Ericsson JL, Eriksson I. Arterial lesions of the radial artery in uraemic patients. Acta Chir Scand. 1979; 145: 415-428.

Newby AC, Zaltsman AB. Molecular mechanisms in intimal hyperplasia. J Pathol. 2000; 190: 300-309. <a href="/cgi/external_ref?access_num=10.1002/(SICI)1096-9896(200002)190:3

Strawn WB, Ferrario CM. Mechanisms linking angiotensin II and atherogenesis. Curr Opin Lipidol. 2002; 13: 505-512.

Boucher P, Gotthardt M, Li WP, Anderson RG, Herz J. LRP: role in vascular wall integrity and protection from atherosclerosis. Science. 2003; 300: 329-332.

Ejerblad S, Ericsson JL. Ultrastructure of the aorta in experimental uraemia. Acta Chir Scand. 1979; 145: 331-343.

Ejerblad S, Eriksson I, Johansson H. Uraemic arterial disease. An experimental study with special reference to the effect of parathyroidectomy. Scand J Urol Nephrol. 1979; 13: 161-169.

Rattazzi M, Bennett BJ, Bea F, Kirk EA, Ricks JL, Speer M, Schwartz SM, Giachelli CM, Rosenfeld ME. Calcification of advanced atherosclerotic lesions in the innominate arteries of ApoE-deficient mice: potential role of chondrocyte-like cells. Arterioscler Thromb Vasc Biol. 2005; 25: 1420-1425.

Giachelli CM, Steitz S. Osteopontin: a versatile regulator of inflammation and biomineralization. Matrix Biol. 2000; 19: 615-622.

Vattikuti R, Towler DA. Osteogenic regulation of vascular calcification: an early perspective. Am J Physiol Endocrinol Metab. 2004; 286: E686-E696.

Matsui Y, Rittling SR, Okamoto H, Inobe M, Jia N, Shimizu T, Akino M, Sugawara T, Morimoto J, Kimura C, Kon S, Denhardt D, Kitabatake A, Uede T. Osteopontin deficiency attenuates atherosclerosis in female apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol. 2003; 23: 1029-1034.

Binder CJ, Chang MK, Shaw PX, Miller YI, Hartvigsen K, Dewan A, Witztum JL. Innate and acquired immunity in atherogenesis. Nat Med. 2002; 8: 1218-1226.

作者单位：Departments of Nephrology (S.B., K.O.), Clinical Biochemistry (S.B., R.B., F.M., L.B.N.), and Pathology (C.B.A.), Rigshospitalet, University of Copenhagen, Denmark.