Maternal Hypercholesterolemia During Pregnancy Promotes Early Atherogenesis in LDL Receptor-Deficient Mice and Alters Aortic Gene Expression Determined by Microarray
Background— Maternal hypercholesterolemia during pregnancy is associated with markedly enhanced fatty streak formation in human fetal aortas and accelerated progression of atherosclerosis in normocholesterolemic children.
Methods and Results— To establish the causal role of maternal hypercholesterolemia in a genetically homogeneous murine model and to test the hypothesis that pathogenic events during fetal development result in persistent changes in arterial gene expression, female LDL receptor-deficient (LDLR−/−) mice were fed regular chow or high-fat diets supplemented with 0.075% or 1.25% cholesterol during pregnancy. Lesion sizes were determined in the aortic origin of their chow-fed offspring at 3 months. Maternal hypercholesterolemia more than doubled lesion sizes in male offspring (P<0.0001 for the 0.0075% cholesterol group). Microarray analysis of the expression of 11 000 murine genes in the nonatherosclerotic descending aorta by Affymetrix gene chips suggested that 139 genes were significantly regulated in offspring of hypercholesterolemic mothers. A subset of 12 of the upregulated transcripts was subjected to secondary analysis by semiquantitative PCR of pooled RNA and 4 genes were found to be upregulated >1.7-fold. Quantitative PCR for one of these genes using RNA from individual mice yielded similar results. Comparative immunostaining for several of the above genes also indicated increased protein content in offspring of hypercholesterolemic mothers.
Conclusions— These findings establish an atherogenic effect of maternal hypercholesterolemia in genetically uniform mice and indicate that changes in aortic gene expression persist long after fetal exposure to hypercholesterolemia. In addition to elucidating pathogenic mechanisms initiated during fetal development, this approach may identify genes in morphologically normal arteries that influence the susceptibility to classical risk factors of atherosclerosis.
Received January 28, 2002; accepted February 5, 2002.
Previous studies established that the atherogenic process in humans may already begin during fetal development and that aortic fatty streak formation at predilection sites of adult atherosclerosis is greatly enhanced in fetuses of hypercholesterolemic mothers.1,2⇓ Although fetal plasma cholesterol levels markedly decrease toward the end of pregnancy1,3⇓ and cholesterol levels during infancy tend to be low, fetal lesions regress only partially and not in all areas of the aorta.4 Instead, the FELIC study (Fate of Early Lesions in Children study, a morphometric postmortem analysis of atherosclerosis in normocholesterolemic children aged 1 to 14 years) showed that maternal hypercholesterolemia was associated with a much faster progression of atherosclerosis. This could not be explained by conventional risk factors of atherosclerosis in mothers or children. A number of theoretical considerations also suggested that differences in the genetic background are not the sole cause, even though genetic traits are likely to contribute to fetal and postnatal atherogenesis.4
Establishing a causal role of maternal/fetal hypercholesterolemia in accelerated fetal or postnatal lesion formation would not only add to our understanding of the pathogenesis of atherosclerosis, but can also have important clinical implications. Given the genetic heterogeneity of human populations and the variability in diets and classical risk factors, unequivocal evidence for a pathogenic role of maternal hypercholesterolemia in fetal lesion formation and enhanced susceptibility to postnatal atherogenesis could only be obtained in genetically more homogeneous experimental models.
A first line of evidence has recently been obtained in NZW rabbits.5,6⇓ In this model, diet-induced maternal hypercholesterolemia was sufficient to enhance lesion formation in their offspring at birth, and both cholesterol-lowering and antioxidant interventions in mothers significantly reduced lesion sizes.5 Maternal hypercholesterolemia also markedly enhanced the postnatal atherogenic response to mild hypercholesterolemia, and treatment of mothers with antioxidants or lipid-lowering drugs effectively prevented this.6 These results establish, in principle, that maternal hypercholesterolemia or the ensuing pathogenic events during fetal development may accelerate atherogenesis later in life and suggest that interventions in hypercholesterolemic mothers during pregnancy may become a novel preventive approach. However, the mechanisms by which in utero programming enhances later atherogenesis are unknown.
The goal of this study was to establish a murine model of maternal hypercholesterolemia and to use it to test the hypothesis that pathogenic events during fetal development influence arterial gene expression later in life. Confirmation of the atherogenic effect of maternal hypercholesterolemia in a second model would greatly strengthen the assumption that maternal hypercholesterolemia contributes significantly to fetal lesion formation in humans. A murine model would also be particularly useful to investigate the underlying pathogenic mechanisms.7 We previously postulated that increased oxidative stress during fetal development would affect the expression of atherogenic factors in the arterial wall and that some of these differences would persist over time.4,6,8,9⇓⇓⇓ This was based on the observation of markedly increased lipid peroxidation in plasma and lesions of mothers, fetuses, and newborns,1,2,5,6⇓⇓⇓ and the fact that multiple signaling pathways are oxidation-sensitive.10–16⇓⇓⇓⇓⇓⇓ The recent availability of murine cDNA microarrays assessing the expression of thousands of genes and expressed sequence tags (ESTs) at a time provides a tool to test this hypothesis and to identify genes that are differentially expressed.17,18⇓
The present study provides evidence that maternal hypercholesterolemia indeed enhances early atherogenesis in genetically uniform mice and that changes in aortic gene expression in their offspring persist well beyond fetal development.
An overview of the experimental design is provided in Figure 1 (top). Three groups of female LDL receptor-deficient (LDLR−/−) mice (10th generation back-bred into C57BL6; from our own breeding colony, established from animals originally obtained from Jackson Labs, Bar Harbor, Me) were fed normal chow or a standard high-fat diet (TD98338, Harlan-Teklad) supplemented with 0.075% or 1.25% cholesterol, beginning 3 to 4 weeks before pregnancy. The 1.25%, but not the 0.075% cholesterol diet, was associated with reduced breeding success and litter sizes. Plasma cholesterol and triglyceride levels were determined in blood samples obtained from the retroorbital plexus, using an automated enzymatic assay (Boehringer Mannheim Diagnostics). Because this requires anesthesia, maternal TC was determined only prior to the start of the diet and immediately before pregnancy, ie, after 3 to 4 weeks on diet. At birth, hypercholesterolemic mothers were switched to regular chow to reduce potential dietary effects during lactation. Offspring were weaned after 3 to 4 weeks and fed regular chow until the age of 3 months. Twelve male or female progeny each of the “control”, “0.075% cholesterol”, or “1.25% cholesterol” groups were used to determine atherogenesis in the aortic origin and gene expression in the aorta. Note that the group designation reflects maternal cholesterol status only, not treatment after birth. Additional mice were used to assess the presence of lesions in the aortic tree.
LDLR−/− mice were used because the cholesterol level of the parent C57BL6 strain cannot be raised to and beyond that seen in human mothers by hypercholesterolemic diets, and because induction of even early fatty streaks in their aortic origin requires dietary supplementation with cholic acid, a proinflammatory agent that may affect genes expression.19 In contrast, in LDLR−/− mice cholate-free diets induce marked hypercholesterolemia and atherogenesis. However, plasma cholesterol levels of 250 mg/dL (expected in offspring fed regular chow) do not induce significant aortic lesion formation at 3 months of age.19–21⇓⇓ Thus, under the present experimental conditions, offspring should have enough lesions in the aortic origin to permit one to assess atherogenic effects of maternal hypercholesterolemia, whereas atherogenesis in the aortic tree should be minimal. Hence, determination of aortic gene expression should not be affected by cellular heterogeneity resulting from the presence of various stages of lesions. Nevertheless, it should be kept in mind that our maternal control group was not normocholesterolemic by murine standards.
Tissue Preparation and Quantification of Atherosclerosis
At 3 months of age, offspring were euthanized by an overdose of anesthetic. Aortas were perfused for 10 minutes under physiological pressure with ice-cold RNAse-free phosphate-buffered saline, pH 7.4, through a canula inserted into the left ventricle. Tissue preparation and quantification of lesion sizes in the aortic origin were performed as described.20,22⇓ Results are reported as mean lesion area per cross-section. The rest of the aorta was used to determine gene expression. Aortas were thoroughly cleaned of adventitial tissue in situ and divided into the arch, thoracic, and abdominal segments. Segments were blot-dried, flash-frozen, and stored in liquid nitrogen until further processing. To assess the absence or presence of lesions in the aortic tree, 2 entire aortas from each group were fixed, paraffin-embedded, and serially dissected from the heart to the iliac bifurcation. Every 10th section was stained with hematoxylin and eosin and examined microscopically.
Determination of Aortic Gene Expression by cDNA Microarray
Total tissue RNA was extracted using an RNeasy kit (QIAGEN) and determined by spectrophotometry. Double-stranded cDNA was synthesized using 50 ng of total RNA and an oligo dT primer with a T7 RNA polymerase promoter site added to its 3′ end, following the protocol recommended by Affymetrix. The cDNA was then used for in vitro transcription in the presence of biotin-labeled ribonucleotides to yield biotin-labeled cRNA (Enzo Diagnostics kit). cRNA was cleaned using Qiagen’s RNeasy mini columns and protocols. RNA integrity was checked by 1% agarose gel electrophoresis and optical density determined at 260 and 280 nm. cRNA (15 μg) was then fragmented and used in a hybridization mix. An aliquot of the mix containing at least 10 μg cRNA was hybridized to the Affymetrix Mu11k subA and subB chips. Chips were then washed, stained, and scanned at the Veterans Affairs Medical Center Microarray Core unit directed by Dr Wachsman. A total of 12 Mu11k chip sets (subA and subB, each) were used to compare normo- and hypercholesterolemic groups and for comparisons between different arterial segments within groups. Pooled tissues of 3 mice each, as well as equally pooled samples of all mice in the 0.075% cholesterol and control groups, were used.
Polymerase Chain Reaction Analysis
Twelve genes exhibiting significant increases in the offspring of hypercholesterolemic mothers by microarray analysis were subjected to secondary analysis by semiquantitative polymerase chain reaction (PCR). For this experiment, the actual mRNA samples of the 0.075% cholesterol group (equal pool from the aortic arch of all 12 mice) and a corresponding sample from 9 control mice were used. Reverse transcription (RT)-PCR was performed by converting 150 ng of total RNA to cDNA using oligo dT priming and SuperScipt RT (Life Technologies) as directed. cDNA was amplified in a 50-μL reaction containing 100 ng of each primer and 1×Master Mix Taq polymerase (Qiagen). Cycle conditions were 94°C, 5 minutes; 45 cycles (94°C, 30 seconds; 58°C, 30 seconds; 72°C, 90 seconds); 72°C, 5 minutes; 4°C. Block temperature was decreased to 20°C and 5-μL aliquots taken after cycles 29, 32, 35, 38, 41, and 45. These were visualized by gel electrophoresis and quantitated using AlphaImager 2000 software (Alpha Innotech Corporation). Primers used were FGF-BP-1, 5′-ggtcagaaagagagccaagaac-3′, 5′-atgtcgcctgtaacatgttgag-3′; FCMo3, 5′-gatgactttcccaacttcatgc-3′, 5′-cttgagaaacagccataggaga-3′; NPAS2, 5′-tcacacacagccatgtcagaac-3′, 5′-agatgttggtgcctgcaaatag-3′; and MERG1, 5′-tcctacaccaatggcatcgaca-3′, 5′-actgggaaacctgagaaagcga-3′. Results were normalized to GAPDH. Due to the relatively high levels of GAPDH in these tissues, the cDNA was diluted 1:50 for the GAPDH measurements. Measurements were repeated 3 times.
Expression of one of the above genes, FCMo3, was also determined by quantitative PCR in all individual samples of the control and 0.075% cholesterol group in which sufficient RNA remained available. Competitor DNA was cloned into TOPO BluntII (Invitrogen) using the same primers as in subsequent PCR: GAPDH, 5′-ccagtatgactccactcacgg-3′, 5′-cagtgggccctcagatgcctg-3′; FCMo3, 5′-gatgactttcccaacttcatgc-3′, 5′-cttgagaaacagccataggaga-3′. To generate the competitor plasmid, 189 base pairs were removed from the GAPDH clone by digestion with AccI and EcoNI, followed by blunt end ligation. Likewise, the FCMo3 clone was digested with BstEII and ligated, removing 261 base pairs. Decreasing amounts of competitor plasmid were added to PCR containing equivalent amounts of cDNA, indicated primer pairs and Taq Master Mix (Qiagen). PCR was carried out by denaturing at 95°C for 5 minutes, followed by 25 (GAPDH) or 35 (FCMo3) cycles: 95°C, 30 seconds; 59°C, 30 seconds; and 72°C, 90 seconds, followed by extension at 72°C for 10 minutes. PCR products were separated on a 1% agarose gel and imaged by UV luminescence.
Gene-chip output files were inspected for artifacts and proper array grid alignment. Data of all experiments were globally normalized using The Equalizer, an application for Windows NT written in Microsoft Visual Basic 6.23 Normalization included linearizing the expression and background correction (98.5%>background). Consistent changes in gene expression due to the high cholesterol diets were identified by nearest neighbors analysis,24 in which the class descriptor was high versus normal cholesterol diet without regard to anatomical location. Genes were prioritized according to the difference between the calculated metric P(g,c) and 100 random permutations of classification. Genes for which significant differences (P<0.05) were detected between control and hypercholesterolemic groups by nearest neighbors analysis and/or Student’s t test were then further analyzed with GeneSpring 3.2.8 (Silicon Genetics). Lesion sizes in the aortic origin, weights and plasma cholesterol and triglyceride data showed normal distribution and comparisons between groups were performed using Student’s unpaired t test, assuming unequal variances. Data shown are mean±SEM.
Sections of the aortic origin were immunostained by an avidin-biotin-alkaline phosphatase (AP) method with a red color substrate21 and counterstained with methyl green. MERG protein was assessed using an affinity-purified rabbit antibody (1:20) to a fusion protein of its human equivalent, HERG, which had a 50/54 amino acid homology to MERG (Alomone Labs). Specificity of staining was verified by competition assays, in which the antibody was preincubated with the antigen. NPAS-2 protein was determined by an affinity-purified rabbit antibody (1 μg/mL) against a 15 amino acid internal peptide sequence of human NPAS-2 (MOP-4) with a 100% homology to mouse NPAS-2 (Alpha Diagnostic International). FGFbp was detected by an antiserum to an affinity-purified recombinant GST-FGFbp fusion protein (1:300 dilution),25,26⇓ a generous gift of Dr A. Wellstein, Georgetown University, Washington, DC.
Both hypercholesterolemic diets markedly raised maternal cholesterol levels (Figure 2A). At the beginning of pregnancy, levels in the 0.075% and 1.25% cholesterol groups were 1063 and 1299 mg/dL, respectively, compared with approximately 250 mg/dL in the control group and all groups prior to intervention (P<0.0001). At birth, mothers of the hypercholesterolemic groups were switched to regular chow and within one week maternal cholesterol levels in all 3 groups were similar (420 to 456 mg/dL). The relative increase in the control group compared with prepregnancy levels was reminiscent of increased cholesterol levels during the third trimester in human mothers.27 Triglyceride levels showed an analogous rise to 120 and 124 mg/dL in the 0.075% and 1.25% cholesterol groups, respectively, compared with 52 mg/dL in the control group (P<0.001) (Figure 2B). Body weights, plasma cholesterol, and triglycerides in male offspring at 3 months old are shown in Table 1. Except for a marginally higher body weight in the 1.25% cholesterol group than in the 0.075% cholesterol group, no significant differences were observed. In contrast, female offspring of the 1.25% cholesterol group had slightly lower body weight and markedly lower plasma cholesterol levels then other female groups (data not shown).
The average size of lesions in sections through the aortic origin of male mice is shown in Figure 3. Compared with offspring of control mice, offspring of mothers fed the 0.075% cholesterol diet showed a >2-fold increase in lesions (P<0.00001). Lesions in the 1.25% cholesterol group were also increased, but the difference failed to reach statistical significance (P=0.11). No significant differences in lesion sizes were seen in female offspring (not shown). Representative examples of lesions in offspring of control and hypercholesterolemic mothers are shown in Figures 1A/1B and 1C, respectively.
In contrast to the aortic origin, serial sectioning of the entire aorta (from the heart to the iliac bifurcation) of 2 additional mice from each group revealed only a few adherent monocytes but no intimal thickening. We therefore bypassed tissue embedding, sectioning, and selective harvesting of arterial cells by laser capture microscope.28 Instead, gene expression was determined in aortic segments containing both the media and endothelial monolayer, using Affymetrix microarrays. Altogether, 6 comparisons were performed between the control group and the 2 hypercholesterolemic groups. Gene expression of 11 000 murine genes/ESTs in all experiments was normalized together and statistical differences between the control and hypercholesterolemic groups were established by nearest neighbors analysis24 and t test. Given that both hypercholesterolemic groups showed increased lesions in the aortic arch, we assumed that genes potentially contributing to atherogenesis should be similarly regulated in both hypercholesterolemic groups, compared with controls. Addition of this requirement reduced the number of significant genes to 139, of which 57 were upregulated and 82 downregulated. All significant genes were then entered into GeneSpring, and a gene tree constructed using a pairwise correlation model. The dendrogram indicating clusters of genes with similar patterns of expression is shown in Figure 1D, together with the color-coded mean expression level of each gene. A graphic representation of all up- and downregulated genes and ESTs is shown in Figure 1E. Interestingly, the vast majority of regulated genes showed a greater degree of regulation in the 0.075% cholesterol group than in the 1.25% group, consistent with more extensive lesions in the aortic origin of the former. Changes in gene expression for most genes were smaller than 3-fold and about two thirds did not exceed 1.5-fold up- or downregulation. All known genes—but not the ESTs—and their GeneBank accession numbers are listed in Table 2.
Twelve genes showing upregulation in the aorta of mice exposed to fetal hypercholesterolemia were subjected to secondary analysis using semiquantitative PCR. The majority of these genes were expressed at relatively low levels, and for 8 genes, differential expression could not be confirmed within the level of experimental error for semiquantitative PCR. However, semiquantitative PCR analysis of 4 of the 12 transcripts indicated 1.7- to 2.9-fold upregulation (Figure 4), in good agreement with the 1.5- to 2.3-fold changes observed for these genes in the microarray experiments. These transcripts are encoded by the fibroblast growth factor binding protein (FGFbp) gene involved in neovascularization and cancer growth,25 the flavin containing monooxidase 3 gene, the NPAS2 gene,29 and the MERG1 potassium channel gene. Quantitative PCR for one of these genes, FCMo3, performed on RNA from individual mice yielded qualitatively similar results (Table 3). Although 82 genes were found to be downregulated at a statistically significant level by microarray analysis, the fold changes were too small to be reliably detected by semiquantitative PCR.
Because aortas were used for RNA extraction, the presence of proteins in experimental mice was assessed in lesion-free sections of the aortic origin. Immunostaining for 3 of the 4 genes found to be upregulated by PCR indicated that the corresponding proteins were also consistently increased in mice of the 0.075% cholesterol group, compared with controls. NPAS-2 protein was expressed in all endothelium of both groups, but staining intensity was generally stronger and some medial cells also stained in the 0.075% cholesterol group (data not shown). Marked differences between groups were also detected for MERG, which was mainly found in some deep medial areas of the aortic origin (Figures 5A through 5D). FGFbp was not significantly expressed in intimal and medial tissues of the control group, but areas of endothelial staining were frequent in the 0.075% cholesterol group (Figures 5E and 5F).
Two main findings emerge from these studies. First, maternal hypercholesterolemia during pregnancy enhances early atherogenesis in a genetically uniform murine model. This considerably strengthens the assumption that the marked increase in fatty streaks seen in human fetuses of hypercholesterolemic mothers is caused, at least in part, by hypercholesterolemia and related pathogenic events during fetal development. The 2-fold increase in lesion sizes in male offspring of hypercholesterolemic mice was smaller than that seen in premature human fetuses and term-born NZW rabbits,1,5,6⇓⇓ suggesting that some enhancement of lesion formation already occurred in our mildly hypercholesterolemic control group compared with truly normocholesterolemic mice. The lack of a significant effect in female offspring contrasts with results in humans and rabbits, but several unrelated studies noted analogous sex differences in mice.22 Nevertheless, the present experiments establish a genetically uniform model in which the mechanisms responsible for fetal atherogenesis and its long-term consequences can be investigated on the molecular level.
The second finding is that maternal hypercholesterolemia during fetal development affects arterial gene expression or transcription, as well as the respective proteins, later in life. Offspring of both groups of hypercholesterolemic mothers showed statistically significant regulation in the expression of 135 genes and ESTs at 3 months old, compared with offspring of normocholesterolemic mothers. The degree of regulation of most of these genes was modest, consistent with the assumption that genes affecting the susceptibility to postnatal atherogenesis exert their influence by acting over prolonged periods, rather than by displaying dramatic levels of regulation. However, the joint analysis of medial and endothelial cells may also have masked selective regulation of endothelial genes. Differences in endothelial staining for the proteins of some of the regulated genes (FGFbp and NPAS2) are consistent with this assumption. Because all offspring were fed regular chow and had virtually identical plasma lipid levels and because the differences were observed in nonatherosclerotic aortas, the regulation of genes cannot be attributed to acute differences in diet-induced oxidative stress or to differences in lesions formed during fetal development.
In contrast to the situation at 3 months, there is ample reason to postulate short-term changes in the expression of many genes during fetal development. Extensive evidence links maternal hypercholesterolemia to enhanced lipid peroxidation during fetal development. This includes the prevalence of oxidation-specific epitopes in fatty streaks of premature human fetuses1 and term-born rabbits,5,6⇓ increased levels of oxidized lipids in plasma and lesions of rabbits, and the reduction of fetal lesion formation by maternal antioxidant treatment.5,6⇓ Increased lipid peroxidation and formation of reactive oxygen species affects several signaling pathways, including the NF-κB pathway,30 FasL and TNF receptor pathways,13–15⇓⇓ and PPARγ,12,31⇓ which regulate expression or transcription of many genes modulating cell recruitment, growth and differentiation, and programmed death.
It can therefore be assumed that multiple genes are up- or downregulated in fetal or neonatal arteries, either as a consequence of acute effects of oxidative stress or secondary to fatty streak formation. What could not be assumed is that pathogenic events during fetal development affect gene expression in morphologically normal aortas much later in life. Microscopic examination of aortas and the lack of regulation of factors traditionally associated with macrophage recruitment, differentiation, or activation suggest that these changes are not the result of differences in intimal lesions. A more likely explanation is that fetal gene modulation results in permanent changes in matrix components, intracellular proteins, enzymes, or DNA. Completely regressed fatty streaks may leave behind similar subtle differences in arterial cells.
Evidence for persistent effects of a brief stimulation during the fetal or neonatal period has been previously reported. For example, enhancement of cholesterol degradation by cholestyramine in neonatal guinea-pigs and very young White Carneau pigeons conveyed a significant protection against dietary hypercholesterolemia in adulthood.32,33⇓ This was associated with an increased level in 7-α-hydroxylase in pretreated adult animals on dietary cholesterol stimulus. Hormonal imprinting in neonatal rats exposed to thyroid stimulating hormone has also been described.34 Interestingly, parental imprinting of the Mas protooncogene also appears to occur during particular stages of fetal development.35
A gene chip-based approach appears particularly attractive to search for evidence of persistent changes in arterial gene regulation. Microarrays have been extensively used for studies of cultured cells, but only recently has their application been extended to cancer36,37⇓ and cardiovascular tissues,38,39⇓ which are greatly complicated by the lack of synchronicity of cell proliferation and differentiation. Tissue heterogeneity may be particularly problematic in atherosclerotic lesions, where macrophages, endothelial cells, smooth muscle cells, and T cells interact. Experimental conditions in this study were chosen to yield descending aortas essentially free of lesions. Indeed, macrophage-specific genes were conspicuously absent, consistent with the lack of foam cells observed in histological sections of the descending aorta. This considerably reduces tissue heterogeneity, but also means that our list was unlikely, a priori, to include many genes expressed in the atherosclerotic intima and contributing to the progression of atherosclerotic lesions. Nevertheless, the list of genes programmed in utero may contain some that influence the susceptibility to postnatal exposure to hypercholesterolemia or other risk factors of atherosclerosis.6
A few of the genes listed in Table 2 are thought to be involved in atherogenesis, such as superoxide dismutase, or neovascularization, such as FGFbp.25 Another gene, NPAS2, has recently been reported to be regulated by nuclear hormone receptors such as RXRα in the vasculature.29 However, most of the genes identified by the microarray experiments have not been previously linked with atherosclerosis.
Although the potential causal role of individual genes remains to be established, the differentially regulated genes identified by these studies provide markers of a “molecular memory” of maternal hypercholesterolemia. Analysis of the mechanisms that control the expression of these genes may provide clues on how hypercholesterolemia or the transient development of fatty streaks results in persistent changes in gene expression or transcription within the artery wall and increased atherogenesis later in life. Comprehensive microarrays containing hybridization targets for all potential murine transcripts should soon be available to facilitate identification of the complete set of genes that are persistently affected by maternal hypercholesterolemia. These arrays may be useful in determining whether patterns of gene expression are normalized in mice lacking candidate effector genes or following maternal treatments that prevent or reduce fetal and postnatal lesion formation, such as cholesterol-lowering or antioxidants.
Insights gained by this the approach may also be of interest with respect to the general question of lesion progression or regression. Both experimental and clinical studies have shown that intense lipid-lowering treatment may result in varying degrees of plaque stabilization or lesion regression. However, to date it is unknown whether such arteries remain more susceptible to future exposure to hypercholesterolemia and therefore require more aggressive preventive measures.
This work was supported by grant HL-56989 (La Jolla Specialized Center of Research in Molecular Medicine and Atherosclerosis). J.S. Welch was supported by a predoctoral training grant from the American Heart Association. We thank Dr William Wachsman for valuable advice on microarrays and their evaluation, Dr Anton Wellstein for generously providing the antibodies to FGFbp, and Florencia Casanada, Mercedes Silvestre, and Joe Juliano for excellent technical assistance.
This article originally appeared Online on February 25, 2002 (Circulation. 2002;105:r19–r26).
- ↵Napoli C, D’Armiento FP, Mancini FP, et al. Fatty streak formation occurs in human fetal aortas and is greatly enhanced by maternal hypercholesterolemia. Intimal accumulation of low density lipoprotein and its oxidation precede monocyte recruitment into early atherosclerotic lesions. J Clin Invest. 1997; 100: 2680–2690.
- ↵Napoli C, Witztum JL, de Nigris F, et al. Intracranial arteries of human fetuses are more resistant to hypercholesterolemia-induced fatty streak formation than extracranial arteries. Circulation. 1999; 99: 2003–2010.
- ↵Napoli C, Witztum JL, Calara F, et al. Maternal hypercholesterolemia enhances atherogenesis in normocholesterolemic rabbits, which is inhibited by antioxidant or lipid-lowering intervention during pregnancy: an experimental model of atherogenic mechanisms in human fetuses. Circ Res. 2000; 87: 946–952.
- ↵Palinski W, D’Armiento FP, Witztum JL, et al. Maternal hypercholesterolemia and treatment during pregnancy influence the long term progression of atherosclerosis in offspring of rabbits. Circ Res. 2001; 89: 991–996.
- ↵Palinski W, Napoli C, Reaven PD. In: Simons DI, Rogers C, eds. Contemporary Cardiology: Vascular Disease and Injury: Preclinical Research. Totowa, NJ: Humana Press; 2000: 149–174.
- ↵Napoli C, Palinski W. Maternal hypercholesterolemia during pregnancy influences the later development of atherosclerosis: clinical and pathogenic implications. Eur Heart J. 2001; 22: 4–9.
- ↵Liao F, Andalibi A, Qiao JH, et al. Genetic evidence for a common pathway mediating oxidative stress, inflammatory gene induction and aortic fatty streak formation in mice. J Clin Invest. 1994; 94: 877–884.
- ↵Ricote M, Huang J, Fajas L, et al. Expression of the peroxisome proliferator-activated receptor gamma (PPARγ) in human atherosclerosis and regulation in macrophages by colony stimulating factors and oxidized low density lipoprotein. Proc Natl Acad Sci U S A. 1998; 95: 7614–7619.
- ↵Dimmeler S, Haendeler J, Galle J, et al. Oxidized low-density lipoprotein induces apoptosis of human endothelial cells by activation of CPP32-like proteases: a mechanistic clue to the ‘response to injury’ hypothesis. Circulation. 1997; 95: 1760–1763.
- ↵Napoli C, Quehenberger O, de Nigris F, et al. Mildly oxidized low density lipoprotein activates multiple apoptotic signaling pathways in human coronary cells. FASEB J. 2000; 4: 1996–2007.
- ↵de Nigris F, Youssef T, Ciafré SA, et al. Evidence for oxidative activation of c-Myc-dependent nuclear signaling in human coronary smooth muscle cells and in early lesions of Watanabe heritable hyperlipidemic rabbits: protective effects of vitamin E. Circulation. 2000; 102: 2111–2117.
- ↵Schena M, Shalon D, Davis R. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science. 1995; 270: 467–470.
- ↵Ishibashi S, Goldstein JL, Brown MS, et al. Massive xanthomatosis and atherosclerosis in cholesterol-fed low density lipoprotein receptor-negative mice. J Clin Invest. 1994; 93: 1885–1893.
- ↵Tangirala RK, Rubin EM, Palinski W. Quantitation of atherosclerosis in murine models: correlation between lesions in the aortic origin and in the entire aorta, and differences in the extent of lesions between sexes in LDL receptor-deficient and apolipoprotein E-deficient mice. J Lipid Res. 1995; 36: 2320–2328.
- ↵Palinski W, Tangirala RK, Miller E, et al. Increased autoantibody titers against epitopes of oxidized LDL in LDL receptor-deficient mice with increased atherosclerosis. Arterioscler Thromb Vasc Biol. 1995; 15: 1569–1576.
- ↵Stuart RO, Bush KT, Nigam SK. Changes in global gene expression patterns during development and maturation of the rat kidney. Proc Natl Acad Sci U S A. 2001; 98: 5649–5654.
- ↵Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science. 1999; 286: 531–537.
- ↵Montes A, Walden CE, Knopp RH, et al. Physiologic and supraphysiologic increases in lipoprotein lipids and apoproteins in late pregnancy and postpartum: possible markers for the diagnosis of “prelipemia”. Arteriosclerosis. 1984; 4: 407–417.
- ↵Bonner RF, Emmert-Buck M, Cole K, et al. Laser capture microdissection: molecular analysis of tissue. Science. 1997; 278: 1481–148327.
- ↵Li JR, Bale LK, Kottke BA. Effect of neonatal modulation of cholesterol homeostasis on subsequent response to cholesterol challenge in adult guinea pig. J Clin Invest. 1980; 65: 1060–1068.
- ↵Subbiah MT, Deitemeyer D, Yunker RL. Decreased atherogenic response to dietary cholesterol in pigeons after stimulation of cholesterol catabolism in early life. J Clin Invest. 1983; 71: 1509–1513.
- ↵Khan J, Bittner ML, Saal LH, et al. cDNA microarrays detect activation of a myogenic transcription program by the PAX3-FKHR fusion oncogene. Proc Natl Acad Sci U S A. 1999; 96: 13264–13269.
- ↵Stanton LW, Garrard LJ, Damm D, et al. Altered patterns of gene expression in response to myocardial infarction. Circ Res. 2000; 86: 939–945.
- ↵Haley KJ, Lilly CM, Yang JH, et al. Overexpression of eotaxin and the CCR3 receptor in human atherosclerosis: using genomic technology to identify a potential novel pathway of vascular inflammation. Circulation. 2000; 102: 2185–2189.