Persistent Inflammatory Pathways Associated with Early Onset Myocardial Infarction in a Medicated Multiethnic Hawaiian Cohort

Abstract

In spite of current standard therapies to target the major pathomechanisms in myocardial infarction (MI), inflammatory gene expression patterns have been consistently revealed in MI patients. In a multiethnic cohort, we aimed to identify MI-associated pathomechanisms that may be unresponsive to medical treatment to improve diagnosis and therapy. Gene expression profiles in whole blood were analyzed in medicated Asian, African American and Caucasian patients living in Hawaii with a history of early MI and age, ethnicity, risk factor and medication-matched controls. PANTHER ontological and Ingenuity Pathway analysis and functional evaluation of the consistently differentially expressed genes identified coordinated up-regulation of genes for inflammation (LGALS3, PTX3, ZBTB32, BCL2L1), T-cell activation (IL12RB1, VAV3, JAG1, CAMP), immune imbalance (IL-8, IL2RA, CCR7, AHNAK), and active atherosclerosis (NR1H4, BIN1, GSTT1, MARCO) that persist in MI patients in spite of concerted treatment efforts to control vascular pathology. Furthermore, significant ethnic differences appear to exist within the active disease mechanisms that need to be further investigated to identify key targets for effective medical intervention.

Keywords

Atherogenesis inflammation immune imbalance microarray myocardial infarction

Background

Cardiovascular disease, most commonly manifested as myocardial infarction (MI), is a significant health problem worldwide. In the United States, MI accounts for one out of every three deaths, and in spite of advances in treatment, remains a pre-eminent health problem.¹

Atherogenesis and formation of atherosclerotic plaques precede MI and numerous efforts have been directed at the development of biomarkers for the diagnosis of atherosclerosis and the atherogenic process.² Recent studies confirmed that circulating blood cells upon encounter of a pathological vascular environment develop a gene expression signature of the vascular disease phenotype.³ Atherosclerosis, coronary artery disease, and MI-associated gene expression patterns have been consistently identified in peripheral blood cell types, including monocytes,⁴ platelets,⁵ and proinflammatory leukocytes that directly contribute to the development of atherosclerosis.⁶ Whole blood gene expression signatures have similarly been confirmed to correlate with the extent of coronary stenosis,⁷ coronary artery disease, and the type and degree of vascular injury, and correctly mirror pathophysiology in the vascular wall.⁸ However, most of the gene expression patterns established in predominantly Caucasian patient groups have not been evaluated in ethnically diverse populations.

Medications commonly used in the prevention and treatment of atherosclerosis and MI include aspirin (ASA), β-adrenergic blockers (β-blocker),⁹ angiotensin converting enzyme inhibitors (ACEI),¹⁰ angiotensin receptor type 1 blockers (ARB)¹¹ and 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins),¹² that target inflammation, cholesterol metabolism and hypertension, the major pathomechanisms, in atherosclerosis and MI. In spite of the wide-spread and long-term use of these therapies, however, circulating blood derived inflammatory gene signatures have been consistently noted in various MI patient cohorts and suggest that certain inflammatory pathways causally relevant to the development of atherosclerosis and MI may be unresponsive to anti-inflammatory medications.⁸ Furthermore, the observed persistent inflammatory pathways while demonstrate common key elements, also reveal considerable diversity in the most significant differentially regulated genes among various MI patient cohorts.

In order to define regulatory pathways associated with MI in an ethnically diverse patient population, and identify inflammatory and atherogenic pathways that may be unresponsive to the currently applied standard therapies for the prevention and treatment of MI, we analyzed gene expression profiles in whole blood samples in a multiethnic patient cohort with a history of early MI (EMI) and compared these to an age, ethnicity, risk factor and medication-matched control cohort (MCON) identified and enrolled through the VA Pacific Islands Healthcare System.

Methods

Study subjects

Following a protocol approved by the VA Pacific Islands Healthcare System (VAPIHCS), University of Hawaii IRBs, and the United States Army Medical Research and Materiel Command Human Research Protection Office, age, gender and ethnicity matched groups of early MI (EMI) and medication-matched control (MCON) subjects were recruited that had no evidence of chronic or active infectious or autoimmune diseases. Using the VAPIHCS data warehouse, eligible subjects were identified based upon the following criteria:

Inclusion criteria

The EMI patient group included individuals that had an ICD-9 code for MI before age 50 for men and 55 for women, that occurred more than 6 months prior to enrollment, and were using any combination of cardiac-related medications including ASA, ACEIs, ARBs, β-blockers and statins. The MCON group included individuals that did not have an ICD-9 code for coronary heart disease, but were taking any combination of the same medications.

Exclusion criteria

1) treatment of active infection within the last month, 2) a clinical history of chronic infection or rheumatologic disease, 3) active use of non-steroidal anti-inflammatory drugs (other than aspirin) for analgesia or inflammatory conditions, 4) a history of an MI within the last 6 months, 5) an absolute monocyte count < 125 cells/μl within the last year, 6) anemia defined as hemoglobin <11 g/dL or HCT < 33% within the last year, 7) CRP ≥ 2.0 mg/L or ESR ≥ 30 mm/hr. Potential subjects with active pregnancy and those with dementia or severe mental illness were excluded based upon review of the medical records.

Eligible subjects were invited for an interview and for informed, consenting subjects a detailed questionnaire was used to record baseline characteristics, including past medical history, compliance with medication use and habits. Routine clinical laboratory testing included a complete blood count with differential, comprehensive metabolic profile, erythrocyte sedimentation rate, C-reactive protein level and in females β-human chorionic gonadotropin level. From selected individuals blood samples were collected for RNA extraction. As no women met the study entry criteria, our recruitment process yielded two age and ethnicity matched study groups with 10 men in each group.

RNA purification and microarray hybridization

A description of the RNA isolation, labeling, double-stranded cDNA synthesis, and hybridization onto Applied Biosystems Human Survey Genome Microarray 2.0 is provided in the Supplementary Material.

Data analysis

Chemiluminescence detection, image acquisition and primary data analysis were performed using the Applied Biosystems 1700 Chemiluminescent Microarray Analyzer. Chemiluminescent images were auto-gridded, spot and spatially normalized, signals were quantified, corrected for background and the final images and feature data processed using the Applied Biosystems 1700 Chemiluminescent Microarray Analyzer software vl.1. A novel algorithm that adaptively thresholds fold change was used to detect a global signal for differential expression between the EMI and MCON sample groups. Data obtained from each chip was first vectorized to form a column of an expression data matrix where the MCON and EMI sample data segregated into two disjoint groups. Each row of the data matrix represented the expression profile of a gene across the EMI and MCON sample groups. The data matrix was quantile normalized to suppress low-frequency systematic error and log2 transformed to equalize variation in fold change over intensity. Box and whiskers plots, pair-wise correlation plots and histograms of log2 expression were generated from raw microarray data to assess data quality. A Ratio-Intensity (RI) scatter plot of mean log2 ratio (MLR) versus mean log2 expression (MLE) was generated where each gene was represented by a point in (MLE, MLR) space. The horizontal MLE-axis was then divided into 51 quantile bins to decouple variation in MLR from MLE where each bin contained approximately 650 genes. Assuming that only a small percentage of genes in a bin were truly differentially expressed, ie, that the signal for differential expression was sparse, a threshold was computed for each bin based on the Donoho-Johnstone universal threshold (DJUT) of wavelet denoising theory^13,14 and Classical Multi-Dimensional Scaling (CMDS or PCA) analysis. The genes that exceeded the DJUT threshold for each bin, together, formed a global estimate of the signal for differential expression.

The differentially expressed genes were analyzed using clustered heatmaps¹⁵ and PCA analysis.¹⁶ Knowledge-based PANTHER (www.pantherdb.org) and Ingenuity's Pathway Analysis (IPA, Ingenuity Systems, Inc, Redwood City, CA) databases were used to identify sets of functionally related genes and signaling pathways significantly overrepresented among genes differentially expressed in EMI samples.

Quantitative real-time PCR

1 μg of total RNA from selected samples was reverse transcribed into cDNA using Superscript III cDNA Synthesis Kit (from Invitrogen, Carlsbad, CA). 100 ng of cDNA template for each sample was amplified with Applied Biosystems TaqMan Gene Expression Assays (Foster City, CA) for the corresponding gene. Gene expression values were calculated relative to GAPDH using the 2^–ΔΔCT method.¹⁷ All assays were done in triplicate and data are reported as mean ± standard error of the mean.

Results

Among medicated early myocardial infarction (EMI) subjects and controls without coronary artery disease or MI on the same medications (MCON), prospective matching ensured close baseline characteristics including body weight, body mass index and rates of cardiac risk factors including hypertension, diabetes, hyperlipidemia, cigarette use, and family history of heart disease (Supplementary Table 1) and laboratory results for clinical parameters (not shown). Blood derived EMI and MCON RNA samples were processed for microarray analysis. Evaluation of the quality of microarray results using box-and-whisker plots (Fig. 1A), pair-wise correlation plots (not shown) and histograms (Fig. 1B) identified sample MCON4 as markedly different and excluded it from further data analysis.

Figure 1.

A) Box and whiskers plots of raw microarray data. Samples are along the horizontal axis while log2 intensity values for each gene are plotted on the vertical axis; the upper and lower ends of the box contain all genes between the first and third quartiles of log2 gene intensity; the red line in each box represents the median log2 intensity value; the upper and lower ends of the dotted vertical lines represent outliers of log2 gene intensity. B) Histogram of raw microarray data, showing the number of genes with a given log2 gene intensity. C) Ratio-intensity (RI) plot of the quantile normalized microarray data. The results of the differential expression analysis for all quantile bins are superimposed on the RI plot. The vertical black and yellow stripes represent 51 sets of approximately 650 genes having similar levels of fluorescent intensity. The dots in red triangles show the up- and those in blue the down-regulated genes.

Differentially expressed genes in EMI

In the EMI samples, 1,203 genes were found differentially expressed (Fig. 1C). The expression value for each differentially expressed gene was plotted in a heatmap to identify groups of genes with consistent similarity in expression across the samples (Supplementary Fig. 1A). Principal component analysis (PCA) of the data clearly separated EMI from MCON individuals (Supplementary Fig. 1B) and identified samples EMI2, EMI5 and EMI7 as distinctly different. Based on PCA results and inspection of clustered heatmaps, 195 genes (Supplementary Fig. 1A, genes 540–735) were excluded from further analysis as their expression variation was contributed by only by two samples (EMI2 and EMI5) and neither subsequent PANTHER nor IPA analysis revealed any significant functional or pathway structure within these genes.

Altered immunity in early MI patients

In order to identify functionally significant EMI-associated patterns among the differentially expressed genes, PANTHER ontological analysis was performed. The biological process of “Immunity and defense” was identified as significantly overrepresented in the set of all differentially expressed genes (P = 1.4E-05) and in the subset of down-regulated genes (P = 1.39E–05) (Supplementary Table 2). Within the subset of down-regulated genes the biological process “Interferon-mediated immunity” (P = 5.72E–03) and the molecular function “Defense/immunity protein” (P = 2.12E–03) were overrepresented. PCA analysis (not shown) and gene distribution within the clustered heatmap (Supplementary Fig. 2) clearly separated EMI from the MCON group suggesting that immunity is a significant distinguishing feature for EMI subjects.

Figure 2.

A) Ingenuity Networks 1–4 (A, C, E, G) and their clustered heatmaps (B, D, F, H). Red color denotes up- and blue down-regulated genes. Solid lines show direct, dotted lines indirect interactions. Arrowheads indicate direction of interactions. Genes with the most connections are hub-genes.

Ingenuity Pathway Analysis (IPA) of the differentially expressed genes identified “Immunological disease and response” and “Immune and lymphatic system development and function” in two of the four most significant IPA networks represented in our expression data (Supplementary Table 3), and additionally, immunity themes within the other two networks. The differentially expressed network genes and those within functional categories (described below) are referenced for their MI-associated roles in Supplementary Table 4.

Multiple active inflammatory networks in medicated MI subjects

In the EMI samples, both IPA Network 1 and 2 (Figs. 2A and 2C) had differential gene regulation pattern with inflammatory functions. The clustered heatmaps (Figs. 2B and 2D) and PCA analysis (not shown) of these networks clearly separated EMI from MCON samples.

Within Network 1, the up-regulated FOS is a highly connected hub gene that promotes proliferation of lymphocytes. The cell fate determinant DACH1 (up-regulated) with a role in immune response is connected to FOS, Apl, and TGF-β. JAG1 (up-regulated) is known to be increased in MI patients and is a ligand for NOTCH-1 that has a role in T-cell maturation and activation. PTX3 (up-regulated) is an acute phase protein that rapidly increases during inflammation with a protective function in cardiovascular disease. CCR7 (down-regulated) is a g-protein coupled receptor chemotactic to leukocytes linked to multiple organ autoimmunity.

In Network 2, CD40LG (down-regulated) is a hub gene expressed on T cells with a central role in atherosclerosis. Elevated CD40LG is a prognostic marker for cardiovascular disease and its down-regulated state in our EMI subjects indicates a response to medications. However, another hub gene, BCL2L1, with an anti-apoptotic and immune activating role, is up-regulated. IL12RB1, also up-regulated, is known to be expressed during chronic inflammation and in MI it contributes to regulation of the cytokine profile of Thl lymphocytes. LGALS3 and CAMP (both up-regulated) induce chemotaxis of inflammatory cells.

Network 3 displays features of immune imbalance, cell survival, chemotactic, and angiogenic functions (Fig. 2E). Its clustered heatmap and PCA analysis (not shown) also differentiates EMI from the MCON group (Fig. 2F). IL-8 (up-regulated) is a hub gene in this network that, through several parallel mechanisms, is a mediator of neutrophil chemotaxis. The down-regulated IL-2RA increases susceptibility for autoreactive T-cells and may indicate an imbalanced immune reaction.¹⁸ The down-regulated WNT1 shows a potential for deficient cell protection and cardiomyocyte survival via PI3K/AKT signaling. WNT5A (up-regulated) has an important role in inflammatory macrophage activation. VEGFB with significant role in neovascularization during inflammation and repair was found down-regulated.

Network 4 is dominated by stress response, cardiac hypertrophy and proinflammatory themes (Fig. 2G) with its clustered heatmap (Fig. 2H) and PCA (not shown) separating EMI from MCON. One of the hub genes, HDAC2 (up-regulated), is an integrator of Hsp70 mediated stress response pathways during heart remodeling in cardiac hypertrophy. BIN1 (down-regulated) has functions in apoptosis and T cell mediated inflammation. TRIO (increased) orchestrates cell-matrix and cytoskeletal rearrangements. SOX6 (decreased) is known to regulate cardiomyocyte development and is a candidate gene for myopathy. IL1RAPL1 (interleukin-1 receptor accessory protein-like-1, down-regulated), belongs to a novel class of the IL-1/Toll receptor family and interacts with neuronal calcium sensor-1 (NCS-1). FREQ (NCS-1) down-regulated in all EMI cases, is a calmodulin-related protein that controls inflammatory reactions.

Immune imbalance, coordinated up-regulation of inflammation and active atherogenesis

In order to gain insight into the regulation of the immunity-related and inflammatory processes in EMI, genes within PANTHER biological process “Immunity and defense” and within IPA Networks 1–4 were plotted on clustered heatmap diagrams. Genes with the most consistent differential expression were identified using clustered heatmaps by two independent evaluators that scored the number of up- or down-regulated genes in the EMI and MCON groups and compared the original expression values.

In EMI, forty-four genes were found most consistently differentially expressed. Among these four major functional categories were identified: 1) inflammation-related, 2) immune imbalance, 3) atherosclerosis, LDL, and cholesterol related genes, and 4) those with a yet unknown function in cardiovascular disease.

Based on expression changes of genes in the inflammation-related category, a pattern of coordinated up-regulation and an overall increase of proinflammatory processes emerged. Proinflammatory genes LGALS3, ROCK1, and PTX3, were consistently up-regulated, while the anti-inflammatory SNFT, and ZBTB32 (PLZP), were consistently down-regulated. Additionally, pro-apoptotic genes NOXA1 and NGFR that would reduce the number of inflammatory cells were consistently down-regulated, while the anti-apoptotic gene BCL2L1 with immune activating effects was consistently up-regulated. Genes that increase T-cell activation, IL12RB1, VAV3, JAG1, and CAMP, were also consistently up-regulated, while SNFT and SIT that decrease T-cell activation were down-regulated. The only proinflammatory molecule that was consistently down-regulated was CD40LG.

In the second functional category, down-regulation of IL2RA, SIT1 and CCR7, and up-regulation of IL12RB1, JAG1, VAV3 and AHNAK with roles T-cell activation indicated immune imbalance.

Within the atherosclerosis category, the anti-atherosclerotic genes, NR1H4, BIN1 and GSTT1, were consistently down-regulated, while the pro-atherosclerotic gene, MARCO, was consistently up-regulated and indicated ongoing atherogenesis.

In addition to genes known to contribute to coronary artery disease, we also identified consistently differentially regulated genes that have not been previously reported in association with MI including HDAC2, WNT-1, TGLN1, DACH1, HLA-DQA1 and SNCA.

Biologically relevant subset of differentially expressed EMI signature genes

Based on biological relevance and the greatest degree of differential expression in EMI patients we have identified 10 out of the 44 most consistently differentially expressed genes for subsequent analysis. A clustered heatmap and PCA analysis (Supplementary Figs. 3A and B) confirmed the ability of these 10 genes to classify the samples as either EMI or MCON. Of these genes we have validated the microarray-derived mean quantile normalized gene expression data (Fig. 3A) for PTX3, LGALS3, CAMP, IL12RB1, JAG1, IL2RA, CCR7, GSTT1, and NGFR using quantitative real time-PCR normalized to GAPDH (Fig. 3B).

Figure 3.

A) Average microarray expression levels (±standard error) and clustered heatmaps of signature genes. Blue hues represent down- and red up-regulation. B) Quantitative real-time PCR validation of the mRNA levels for the same genes. Expression values were calculated relative to GAPDH levels, assays were performed in triplicate and data are shown as mean ± standard error of the mean.

Discussion

In this study of peripheral blood cell gene expression profiling in a prospectively matched cohort of Asian, African-American and Caucasian patients living in Hawaii, diagnosed with early myocardial infarction, and undergoing standard post-MI therapy, integrated functional analysis of the differentially expressed genes identified proinflammatory gene signatures. CD40LG that is actively involved in the pathophysiology of coronary artery syndromes was down-regulated in response to therapy.¹⁹ However, prominent signatures for active inflammation involving FOS, DACH1, PTX3 and JAG1 were still present. Inflammatory functions were also represented by disregulated IL 12RB1, LGALS3, CAMP, and BCL2L1, an immune imbalance pattern included IL-8 and the IL-2RA/IL-2 system, and additionally WNT1 disregulation indicated likely compromised cardiomyocyte survival. The hub gene HDAC2 revealed an overall stress response, and we also noted gene signatures consistent with cardiac hypertrophy and myopathy.

Functional analysis of the 44 most consistently differentially expressed genes further confirmed immune imbalance, coordinated differential regulation of inflammatory pathways and active atherosclerosis as the persisting main disease mechanisms in medicated EMI patients. Of these genes the proinflammatory LGALS3 and PTX3 and the T-cell activating IL12RB1, JAG1 and CAMP were consistently up-regulated, while the pro-apoptotic, anti-inflammatory NGFR, auto-immunity protective CCR7, IL2RA and the anti-atherosclerotic GSST1 consistently down-regulated. Altered peripheral blood expression of a similarly small number²⁰ of genes have been shown to be a signature significantly associated with the severity of coronary artery disease⁷ and predictive for the severity of atherosclerosis.⁸ The novel, differentially expressed and early MI-associated genes that we have identified include HDAC2 an integrator of stress response pathways during heart remodeling, WNT1 with a known role in cardiomyocyte survival, TGLN1 (TAGLN) an actin-binding SM22/transgelin-related protein, DACH1 a cell fate determinant, SNCA linked to multiple system atrophy, and HLA-DQA1 a regulator of immune-specific response to anti-myocardium antibodies.

The short lists of differentially expressed genes in our multiethnic EMI cohort proved different from the signature genes reported for coronary artery disease in Caucasian cohorts.^7,8 Cardiovascular disease susceptibility and SNP frequencies in cardiovascular disease genes are known to vary among populations.²¹ In a recent study, the biological processes of apoptosis, lipoprotein metabolism and immunity showed the greatest difference among Africans, Europeans and South Asians.²² Significant differences in allele frequencies were noted for Toll-like receptors,²³ immune response involving IL-2, IL-6, and IL-10,^24,25 and apoptosis including BCL2L1, IL-4, and IL-6 among Blacks, Whites, and Asians.²² In Japanese, Taiwanese²⁶ and Korean populations, elevated BRCAl-associated protein, BRAP, was shown to contribute to increased MI risk.²⁷ These observations are in agreement with the differential representation in our EMI cohort of a IL-1/Toll receptor family member, IL1RAPL1, numerous interleukins and IL-receptors, BCL2L1, BRCA1 and the immune response regulator HLA-DQA1 that was recently reported in association with cardiomyopathy in Chinese patients.

Chronic local and systemic inflammation and immune disregulation²⁸ play prominent roles in the pathogenesis of MI. However, given that our EMI patients have been under long-term anti-inflammatory and anti-atherosclerotic therapy, the robust representation of inflammation, immune imbalance, and active atherogenesis gene expression signatures was unanticipated. In a recent study of coronary artery stenosis patients with a comparable percentage of subjects receiving ASA, ACEIs, β-blockers and statins, differential blood expression pattern of 160 genes also revealed altered regulation of immunity, persistent inflammation and apoptosis.⁸ The anti-atherosclerotic and anti-inflammatory statins are one of the most important medications in cardiovascular diseases that favorably affect platelet adhesion, endothelial function, plaque stability and cholesterol synthesis, and were thought to also inhibit TNF-α, IL-1 and IL-8.²⁹ However, emerging evidence revealed that statins did not have significant effect on markers of inflammation and platelet activation,³⁰ release of IL-1β and IL-6, and that actually increased the release of TNF-α and IL-8. Our data, particularly the robust representation of IL-8-associated inflammation in medicated EMI patients, support these observations.

Conclusion

Our results demonstrate that active pathogenesis including multiple inflammatory pathways, immune imbalance, stress response and atherogenesis persists in EMI patients in spite of concerted treatment efforts to protect these subjects from continued vascular pathology. Furthermore, while persistent inflammation, undetected by standard laboratory tests, appears to be the most significant ongoing pathomechanism in this cohort, ethnic differences exist among Asian, African American and Caucasian MI patients in the active inflammatory pathways. Further ethnic cohort-based studies are needed, therefore, to determine the critical ethnicity-specific pathology profiles, identify optimal candidate targets and develop effective treatment modalities for post-MI patients with a focus on inflammatory, immune, and atherogenic pathways that are unresponsive to current therapies.

Disclosures

This manuscript has been read and approved by all authors. This paper is unique and not under consideration by any other publication and has not been published elsewhere. The authors and peer reviewers report no conflicts of interest. The authors confirm that they have permission to reproduce any copyrighted material.

Footnotes

Acknowledgments

We are thankful to Greg Kronmal at Applied Biosystems for his technical assistance and insightful discussions. This work was supported by the Department of Defense Grant W81XWH 06 2 0068 USAMRAA and resources and the use of facilities at the Spark Matsunaga VA Medical Center, Honolulu, Hawaii. The contents of this article do not necessarily reflect the views of the Department of Veterans Affairs or the United States Government.

Supplementary Material

References

Lloyd-Jones

, Adams

R.J.

, Brown

T.M.

. Executive summary: heart disease and stroke statistics—2010 update: a report from the American Heart Association. Circulation. 2010; 121: 948–54.

Rosenson

R.S.

, Koenig

Utility of inflammatory markers in the management of coronary artery disease.

Am J Cardiol. 2003; 92: 10i–8.

Aziz

, Zaas

, Ginsburg

G.S.

Peripheral blood gene expression profiling for cardiovascular disease assessment.

Genomic Med. 2007; 1: 105–12.

Kang

J.G.

, Patino

W.D.

, Matoba

, Hwang

P.M.

Genomic analysis of circulating cells: a window into atherosclerosis.

Trends Cardiovasc Med. 2006; 16: 163–8.

Healy

A.M.

, Pickard

M.D.

, Pradhan

A.D.

. Platelet expression profiling and clinical validation of myeloid-related protein-14 as a novel determinant of cardiovascular events. Circulation. 2006; 113: 2278–84.

Libby

Inflammation in atherosclerosis.

Nature. 2002; 420: 868–74.

Wingrove

J.A.

, Daniels

S.E.

, Sehnert

A.J.

. Correlation of peripheral-blood gene expression with the extent of coronary artery stenosis. Circ Cardiovasc Genet. 2008; 1: 31–8.

Sinnaeve

P.R.

, Donahue

M.P.

, Grass

. Gene expression patterns in peripheral blood correlate with the extent of coronary artery disease. PLoS One. 2009; 4: e7037.

Beattie

M.S.

, Shlipak

M.G.

, Liu

, Browner

W.S.

, Schiller

N.B.

, Whooley

M.A.

C-reactive protein and ischemia in users and nonusers of beta-blockers and statins: data from the Heart and Soul Study.

Circulation. 2003; 107: 245–50.

10.

Di Napoli

, Papa

Angiotensin-converting enzyme inhibitor use is associated with reduced plasma concentration of C-reactive protein in patients with first-ever ischemic stroke.

Stroke. 2003; 34: 2922–9.

11.

Kranzhofer

, Schmidt

, Pfeiffer

C.A.

, Hagl

, Libby

, Kubler

Angiotensin induces inflammatory activation of human vascular smooth muscle cells.

Arterioscler Thromb Vasc Biol. 1999; 19: 1623–9.

12.

Kwak

, Mulhaupt

, Myit

, Mach

Statins as a newly recognized type of immunomodulator.

Nat Med. 2000; 6: 1399–402.

13.

Donoho

, Johnstone

Ideal spatial adaptation by wavelet shrinkage.

Biometrika. 1994; 81: 425–55.

14.

Sabatti

, Karsten

S.L.

, Geschwind

D.H.

Thresholding rules for recovering a sparse signal from microarray experiments.

Math Biosci. 2002; 176: 17–34.

15.

Calvano

S.E.

, Xiao

, Richards

D.R.

. A network-based analysis of systemic inflammation in humans. Nature. 2005; 437: 1032–7.

16.

Borg

I.G.P.

Modern multidimensional scaling: theory and applications. 2nd ed. Springer Verlag: New York; 2005.

17.

Livak

K.J.

, Schmittgen

T.D.

Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.

Methods. 2001; 25: 402–8.

18.

Watford

W.T.

, Moriguchi

, Morinobu

, O'shea

J.J.

The biology of IL-12: coordinating innate and adaptive immune responses.

Cytokine Growth Factor Rev. 2003; 14: 361–8.

19.

Heeschen

, Dimmeler

, Hamm

C.W.

. Soluble CD40 ligand in acute coronary syndromes. N Engl J Med. 2003; 348: 1104–11.

20.

, Liew

C.C.

Gene profiling identifies secreted protein transcripts from peripheral blood cells in coronary artery disease.

J Mol Cell Cardiol. 2003; 35: 993–8.

21.

Lanfear

D.E.

, Marsh

, Cresci

, Shannon

W.D.

, Spertus

J.A.

, Mcleod

H.L.

Genotypes associated with myocardial infarction risk are more common in African Americans than in European Americans.

J Am Coll Cardiol. 2004; 44: 165–7.

22.

Kullo

I.J.

, Ding

Patterns of population differentiation of candidate genes for cardiovascular disease.

BMC Genet. 2007; 8: 48.

23.

Ryan

A.W.

, Mapp

, Moyna

. Levels of interpopulation differentiation among different functional classes of immunologically important genes. Genes Immun. 2006; 7: 179–83.

24.

Hoffmann

S.C.

, Stanley

E.M.

, Cox

E.D.

. Ethnicity greatly influences cytokine gene polymorphism distribution. Am J Transplant. 2002; 2: 560–7.

25.

Ozaki

, Sato

, Inoue

. SNPs in BRAP associated with risk of myocardial infarction in Asian populations. Nat Genet. 2009; 41: 329–33.

26.

Hinohara

, Ohtani

, Nakajima

. Validation of eight genetic risk factors in East Asian populations replicated the association of BRAP with coronary artery disease. J Hum Genet. 2009; 54: 642–6.

27.

Liang

K.P.

, Kremers

H.M.

, Crowson

C.S.

. Autoantibodies and the risk of cardiovascular events. J Rheumatol. 2009; 36: 2462–9.

28.

Biasucci

L.M.

, Biasillo

, Stefanelli

Inflammatory markers, cholesterol and statins: pathophysiological role and clinical importance.

Clin Chem Lab Med. 2010; 48: 1685–91.

29.

Fichtenbaum

C.J.

, Yeh

T.M.

, Evans

S.R.

, Aberg

J.A.

Treatment with pravastatin and fenofibrate improves atherogenic lipid profiles but not inflammatory markers in ACTG 5087.

J Clin Lipidol. 2010; 4: 279–87.

30.

Nilsson

, Eriksson

, Cherfan

, Jonasson

Effects of Simvastatin on Proinflammatory Cytokines and Matrix Metalloproteinases in Hypercholesterolemic Individuals.

Inflammation. 2010.