View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Journal of the American College of Cardiology Vol. 50, No. 12, 2007 © 2007 by the American College of Cardiology Foundation ISSN 0735-1097/07/$32.00 Published by Elsevier Inc. doi:10.1016/j.jacc.2007.06.012

STATE-OF-THE-ART PAPER

C-Reactive Gene Polymorphisms, C-Reactive Protein Levels, and Cardiovascular Disease Risk

Fadi G. Hage, MD, Alexander J. Szalai, PHD Birmingham, Alabama

C-reactive protein (CRP), a blood marker of inflammation and a hallmark of the acute-phase response, has been shown to be a powerful and specific predictor of cardiovascular event risk in populations of otherwise healthy persons. Here we review what is known about CRP gene polymorphisms, discuss how these might affect the epidemiology of CRP and our understanding of CRP’s contribution to cardiovascular disease, and examine their potential clinical usefulness. Evidence shows that certain subtle variations in the CRP gene sequence, mostly single nucleotide polymorphisms, predictably and strongly influence the blood level of CRP. Some of these varia- tions are associated with clinical correlates of cardiovascular disease. If future studies can establish with cer- tainty that CRP influences cardiovascular biology, then CRP gene profiling could have clinical utility. (J Am Coll Cardiol 2007;50:1115–22) © 2007 by the American College of Cardiology Foundation

C-reactive protein (CRP) was discovered in 1930 by Tillet “high-sensitivity CRP” assays, technologies that allow cli- and Francis during their studies of patients with acute nicians to detect the low levels of blood CRP in ostensibly pneumonia (1). They found that when from febrile healthy people, has resulted in accumulation of vast patients was mixed with a cell-wall component of pneumo- amounts of data linking blood CRP to various kinds of cocci that they called “Fraction C,” a precipitate formed. cardiovascular diseases (5). At the same time, new evidence This property was subsequently found to be due to reactivity from animal models indicates that CRP might actually have of CRP, present at high levels in “acute-phase” sera, with a a pathogenic role in vascular disease (6–9). polysaccharide in the pneumococcal cell wall (C- It has long been recognized that environmental variables polysaccharide). C-reactive protein was the first of many and patient behaviors and traits such as smoking, infections, acute-phase reactants subsequently discovered (2), but today age, gender, levels, and blood pressure can contribute elevation of CRP is still considered the hallmark of the to variation in baseline CRP level1 (10). Indeed, obesity is a acute-phase response. major determinant of CRP levels in humans, and elevated C-reactive protein is a member of the phylogentically levels of CRP predict the development of type 2 diabetes ancient and conserved pentraxin family of . Human mellitus and the metabolic syndrome (11,12). Elevated CRP (Fig. 1) consists of 5 noncovalently bound subunits of CRP levels in these conditions are related to insulin resis- 206 amino acids, each arranged symmetrically around a tance (13), and weight loss can reduce CRP levels (14). central pore. The molecule has a ligand recognition face that Newer evidence indicates an additional and substantial ϩ contains a Ca2 -dependent binding site, as well as an genetic component (10,15,16). The realization that CRP effector molecule binding face that is able to initiate fluid- genetic polymorphisms do exist and that certain of these phase pathways of host defense (by activating the comple- directly and predictably influence steady-state blood CRP ment system) and cell-mediated pathways (by activating level could be of substantial clinical importance, because complement and by binding to the Fc receptors of immu- genetic predisposition to high baseline CRP might account noglobulin G) (3)(Fig. 1). The rise in blood CRP after for a significant proportion of people with a higher than tissue insult or injury is rapid and robust, with levels average risk of heart disease (17). Herein we review the increasing by as much as 1,000-fold above baseline within available evidence showing that a variety of CRP gene 24 h. This behavior makes blood CRP an ideal clinical polymorphisms are statistically associated with blood CRP marker of a patient’s general health status, and it has thus level, that certain of these are biologically functional in that been used for decades (4). The recent introduction of they directly alter CRP blood levels, and that some CRP variants are linked to a risk of heart and blood-vessel disease.

From the Department of Medicine, University of Alabama at Birmingham, Birming- ham, Alabama. 1Because of the highly skewed distribution of human CRP blood levels, the Manuscript received March 14, 2007, revised manuscript received May 7, 2007, population statistic most often reported in the epidemiologic literature is the median. accepted June 5, 2007. Mean CRP levels and transformed CRP levels are also reported. 1116 Hage and Szalai JACC Vol. 50, No. 12, 2007 CRP Gene Polymorphisms September 18, 2007:1115–22

Abbreviations Localization and and Acronyms Structure of the CRP Gene coronary artery On the basis of the amino acid ؍ CABG bypass graft sequence of the protein, White- C-reactive protein head et al. (18) synthesized ؍ CRP interleukin CRP-specific oligonucleotides ؍ IL systemic lupus and used these to screen a human ؍ SLE erythematosus c-deoxyribonucleic acid li- single nucleotide brary. The clone pCRP1 was ؍ SNP polymorphism thus isolated, sequenced, and used as a probe to localize the human CRP gene to chromosome 1. This launched the Figure 2 The CRP Gene Is Located on Chromosome 1 era of CRP genetics. C-reactive protein was subsequently mapped to the proximal long arm of chromosome 1 in the The gene per se is comprised of 2 exons (red boxes) separated by a single intron encoding a dinucleotide (gt) repeat. Exon 1 encodes an 18 amino acid 1q23.2 region (19,20)(Fig. 2). The CRP gene sequence leader peptide (green box) and the first 2 amino acids of the mature protein. was simultaneously determined in 1985 by 2 different The cross-hatched box demarcates the position of a reported alternative tran- groups (21,22), both reporting that it is composed of 1 script. CRP ϭ C-reactive protein; UTR ϭ untranslated regions (yellow boxes). intron separating 2 exons (Fig. 2). The first exon encodes a signal peptide and the first 2 amino acids of the mature protein. This is followed by a 278-nucleotide-long intron cleic acid production. Several reviews of that subject are that includes a GT repeat sequence. The second exon available (3,24,25). Fundamentally, regulation of CRP encodes the remaining 204 amino acids, followed by a expression occurs mostly at the transcriptional level, with stop codon. Goldman et al. (23) were the first to show interleukin (IL)-6 being the major inducer and IL-1 that the GT stretch in the intron is polymorphic in acting synergistically to enhance the effect (26). Accord- length. ingly, certain polymorphisms of the IL-1 (27,28) and Once the CRP gene was identified and its sequence IL-6 (29) genes associate with differences in blood CRP reported, several groups investigated its proximal pro- level. Both IL-1 and IL-6 polymorphisms have been linked moter region and how this regulates messenger ribonu- to myocardial infarctions and stroke (30,31) and to mortal- ity after acute coronary syndromes (32). Studies using mice expressing human CRP from the human CRP promoter have shown that although IL-6 is necessary for the induc- tion of the CRP gene, it is not by itself sufficient (33,34). In humans, CRP blood levels are generally higher in females than in males (35), whereas in mice, expression of the human CRP transgene exhibits the opposite pattern (33). It is important to note that in vitro studies of regulation of CRP gene expression have focused solely on primary hepa- tocytes, hepatocyte cell lines, or a variety of transfected cell lines (3,24,25). With the recent realization that CRP can be locally produced by the vasculature (7,36), however, this area needs to be reinvestigated to determine whether the same regulatory mechanisms are operating in cells of non- hepatic origin. In this review we will not discuss further IL-6, IL-1, gender , or any other of the multitude of trans- acting factors and their genes that might themselves be polymorphic and might participate in CRP gene regulation. Rather, we will focus on the CRP gene per se and systematically discuss its many polymorphic cis-acting ele- Figure 1 Blood CRP Circulates as a Pentamer ments. We will in turn examine evidence that the CRP gene One face of the pentamer supports multipoint attachment to ligands and coding and promoter region is polymorphic, that this ligand-decorated surfaces; the other face binds C1q and Fc␥R. Only some of sequence variation affects how much blood CRP is made, the many C-reactive protein (CRP) ligands are listed. HDL ϭ high-density ; oxLDL ϭ oxidized low-density lipoprotein. and that this in turn affects CRP’s association to cardiovas- cular disease. JACC Vol. 50, No. 12, 2007 Hage and Szalai 1117 September 18, 2007:1115–22 CRP Gene Polymorphisms

SNP in the CRP Gene Table 1 SNP in the CRP Gene

Nucleotide Associated SNP Position Commonly Relative Allele With Higher Gene Region Reported in the Literature NCBI refSNP ID Frequency* CRP Blood Level Association Studies† Promoter Ϫ757 rs3093059 TϾCT (47,58,60,72) Ϫ717 rs2794521 TϾCT(46,49,52,57,58,66,67) Ϫ409 rs3093062 GϾAG (51) Ϫ390 rs3091244 CϾTϾAT(16,51,52,55,67,68,70,71) Intron ϩ29 rs1417938 AϾTA(49,50,60,67,72) Exon 2 ϩ1059 rs1800947 GϾCG(50,52,60–63,65,67,71,72) 3= untranslated ϩ219 rs3093066 CϾAC (49) ϩ1444 rs1130864 CϾTT(46–48,52,67,69,71) 3= flanking ϩ1846 rs1205 CϾTC (65–67,69,72) ϩ2911 rs3093068 CϾGG (69)

*Alleles are reported in order from most frequent to least frequent as listed on the National Center for Biotechnology Information (NCBI) website. Depending on the deoxyribonucleic acid strand sequenced, allele designations may differ from those commonly reported in the literature. †References to articles discussed in the text that studied the association of the indicated single nucleotide polymorphism (SNP) to C-reactive protein (CRP) blood level or disease status. refSNP ϭ reference single nucleotide polymorphism.

The CRP Gene Is Polymorphic CRP Gene Polymorphisms Associate With Differences in CRP Blood Levels Pankow et al. (37) estimated that the interindividual varia- tion in blood CRP level is 35% to 40% heritable. This has Ours was the first group to report an association between since been confirmed in twin studies (38–40). For example, CRP gene polymorphism and baseline blood CRP (45). We MacGregor et al. (39) compared monozygotic versus dizy- examined the relationship between GT repeat alleles and gotic twin pairs and found that, although CRP was associ- blood protein levels in patients with systemic lupus ery- ated with body mass index, smoking status, and thematosus (SLE) versus healthy control patients and replacement therapy, there was a greater correlation of CRP therein identified 13 alleles ranging in length from 9 to 25 9 25 16 21 level among monozygotic than dizygotic twins, with an repeats (GT to GT ). Two of these, GT and GT , estimated heritability of 52%. Similarly, Retterstol et al. (40) accounted for more than 75% of all the alleles found. The found in monozygotic twins the within-pair correlation statistical relationship between CRP genotype and CRP blood level was best described by a sine-wave model having coefficient for CRP level was 0.40. Soon after it was 16 21 established that differences in CRP level associate with a periodicity of 5 GT repeats, with the GT and GT variants associated with lowest CRP levels (Fig. 3). No genetic differences, efforts were made to identify different particular GT allele was associated with SLE (45). CRP gene polymorphisms to determine whether they di- Soon after our study was published, Brull et al. (46) rectly affect the protein’s blood level. We have already reported their own findings. They measured blood CRP in mentioned that soon after the gene was sequenced, a healthy men before and after 48 h of endurance exercise, and polymorphism in the length of the GT repeat in the CRP in patients before and after coronary artery bypass graft intron was identified (23,41). Some time later, Cao and (CABG) surgery and screened each of their respective CRP Hegele reported (42) the first coding region variation, i.e., a genes. Two new polymorphisms were identified; a single nucleotide polymorphism (SNP) in exon II of CRP. Ϫ717G/A SNP (rs2794521) in the CRP promoter and a ϩ 2 That SNP, 1059 G/C (rs1800947) (Table 1), is silent at ϩ1444C/T SNP (rs1130864) in the 3= untranslated region. ¡ ¡ the amino acid level (CTG CTC, Leu Leu), and the The former did not associate with CRP, but ϩ1444TT rarer C allele was reportedly found in Caucasians but not homozygotes had significantly higher CRP at baseline, after Inuits (43). Pronounced ethnic and race effects are now exercise training, and after CABG. The ϩ1444C/T SNP by known to be a general characteristic of the distribution of itself accounted for 3.7% of the total variance in baseline CRP gene polymorphs. Indeed, a recent systematic re- blood CRP. Later the same group reported that CRP was sequencing of the CRP gene revealed as many as 40 SNPs higher in periodontal patients carrying the ϩ1444T versus forming as many as 29 different haplotypes, with by far the the C allele (47), and another group showed that healthy highest nucleotide diversity observed in African Americans young men with the ϩ1444TT genotype had 64% higher (44). It is indisputable that the CRP gene is polymorphic. blood CRP than those with the ϩ1444CC genotype (48). Many other CRP polymorphisms with association to blood CRP level have since been reported. Thus, Obisesan et al. (49) identified association between a Ϫ717A/G SNP 2Throughout the text, SNPs are identified using the terminology most commonly used in the literature. Wherever possible each is also cross-referenced to the NCBI and CRP levels before and after exercise training: blood database using the reference (ref)SNP ID number or ‘rs’ number. CRP was higher in Ϫ717AA homozygotes. A similar 1118 Hage and Szalai JACC Vol. 50, No. 12, 2007 CRP Gene Polymorphisms September 18, 2007:1115–22

values in individuals carrying Ϫ390TT. Both SNPs reside within E-box elements we named E-box 1 and E-box 2, respectively. Using electrophoretic mobility shift assays, we proved that both E-boxes are functional and that transcrip- tion factors are capable of binding E-box 1. Importantly, the SNP in E-box 1 affects its ability to interact with transcrip- tion factors. We also showed that individuals with a Ϫ409G/Ϫ390T haplotype (the only haplotype found that encoded a high transcription factor binding version of E-box 1 and a transcription factor binding version of E-box 2) had highest blood CRP, whereas individuals with Ϫ409A/Ϫ390T (weak transcription factor binding versions of both E-boxes) had the lowest (Fig. 4A). Individuals with either of the 2 remaining haplotypes had intermediate CRP levels. Finally, using CRP promoter-firefly luciferase report- ers, we confirmed that transcription at baseline and in response to IL-6 was different for the different E-box haplotypes (Fig. 4B). Impressively, carriers of the Ϫ409G/ Ϫ Figure 3 Association Between Baseline CRP and 390T haplotype who never smoked still had higher CRP the GT Repeat Polymorphism in the CRP Gene levels than Ϫ409G/Ϫ390C carriers who currently did (51).

Each circle and whisker is the mean Ϯ SEM for baseline (age-adjusted) CRP Kovacs et al. (52) also identified and studied the triallelic obtained from the indicated number of individuals. The correlation coefficient Ϫ390C/T/A SNP. They examined the correlation between (r ) for the sine-wave regressions is given. The arrows point to the common GT16 and GT21 variants having the lowest (approximately equal) CRP levels. CRP ϭ C-reactive protein. correlation was reported for a ϩ219G/A SNP (rs3093066) in the 3= untranslated region: GG homozygotes had higher CRP levels at baseline and after exercise (49). Impressively, in a multivariate analysis, the Ϫ717A/G and ϩ219G/A SNPs together accounted for almost 9% of the total varia- tion in CRP blood level (49). Finally, a ϩ29A/T SNP (rs1417938) was found to be significantly associated with baseline CRP level: individuals with the A allele had higher blood CRP, a relationship that persisted even after control- ling for clinical characteristics (50). It is now clear that polymorphisms in the CRP gene associate with differences in CRP blood level.

Some CRP Gene Polymorphisms Are Functional

None of the polymorphisms so far discussed has been shown to be functional, that is, to directly contribute to differences in baseline CRP blood level among individuals. Few studies have actually pursued this question, and in the absence of evidence, the common assumption made is that CRP polymorphisms alter the structure of neighboring deoxyri- bonucleic acid and thus impede or bolster its translation to messenger ribonucleic acid, or that known polymorphisms are in linkage disequilibrium with yet-to-be identified Figure 4 Association of CRP Promoter functional polymorphisms. To investigate this issue, we Haplotype With Baseline CRP sequenced the CRP promoter in 287 healthy individuals Baseline C-reactive protein (CRP) concentration for the indicated number of and identified 2 SNPs, a biallelic one at nucleotide individuals having the various haplotypes (A) and the corresponding relative Ϫ409G/A (rs3093062) and a triallelic one at Ϫ390C/T/A luciferase light units (the ratio of firefly vs. Renilla luciferase activity) (B) is reported as the mean ϩ SEM (bars and whiskers). The numbers above the (rs3091244) (51). Blood CRP was lowest in people with the bars give the fold-increase in promoter activity relative to the steady state. Ϫ409AA genotype, intermediate in Ϫ409GA, and highest IL ϭ interleukin. Modified with permission from Szalai et al. (51). in Ϫ409GG, and there was a tendency toward high CRP JACC Vol. 50, No. 12, 2007 Hage and Szalai 1119 September 18, 2007:1115–22 CRP Gene Polymorphisms this SNP, a biallelic one at Ϫ717A/G, the ϩ1059G/C SNP the C allele, and there was a higher frequency of ϩ1059C in in exon 2, and the ϩ1444C/T SNP in the 3= flanking patients with a history of acute myocardial infarction com- region. That study confirmed that the triallelic SNP corre- pared with healthy control patients. In a nested-case control lated with baseline CRP. In the large Framingham Heart study of SLE (64), certain alleles of the intronic GT repeat Study (16), 9 of 13 CRP SNPs were found to be associated were found to be associated with vascular events (myocardial with blood CRP after adjustment for clinical covariates. Of infarction, stroke, peripheral arterial thrombosis, CABG, these 9, only the triallelic SNP remained associated with angina or claudication). Systemic lupus erythematosus pa- blood CRP after accounting for correlation among SNPs. tients with these vascular events were more likely than SLE Finally, because CRP level after an acute coronary syndrome patients without them to have blood CRP in the highest is predictive of recurrent events (53,54), Danik et al. (55) quintile of measured values, and these patients had a higher evaluated patients with acute coronary syndrome and found prevalence of the GT20 variant (64). In another study of that the highest CRP blood level was associated with a SLE patients, Russell et al. (65) directly sequenced the CRP Ϫ757T/C SNP (rs3093059) and again with the Ϫ390C/ gene in British families and found a correlation between T/A SNP. The evidence from our initial study (51) and baseline blood CRP, the ϩ1059G/C SNP, and a independent additional studies thus points to at least 1 CRP ϩ1846G/A SNP (rs1205) in the 3= flanking region. The gene polymorphism that might truly be functional: Ϫ390C/ latter was associated with SLE and the presence of anti- T/A. nuclear . Other efforts to associate CRP gene polymorphism with CRP Gene Polymorphism blood CRP level and disease status are noteworthy. For Associates With Disease Status example, in Han Chinese subjects, although neither Ϫ717A/G nor ϩ1846 A/G was found to correlate with Some of the most compelling reports of a strong association blood CRP, the frequency of the Ϫ717A allele was signif- between CRP polymorphism and disease status did not icantly higher among patients with coronary heart disease include measurement of blood CRP level as a variable. For than control patients. Impressively, in that study, individu- example, Roy et al. (56) linked one poly(GT) allele (likely als carrying the A allele had an odds ratio for coronary heart GT16)(Fig. 3) to increased susceptibility to infection with disease of 6.8 compared with those not carrying that allele Streptococcus pneumonia, and Wolford et al. (57) found that (66). This finding has since been confirmed, as the same Ϫ717TT was associated with an increased prevalence of SNP was found to correlate with occurrence of myocardial type 2 diabetes mellitus. More recently, it was reported (58) infarction or thromboembolic stroke in the Physician’s that having the C allele at the Ϫ717T/C locus was signif- Health Study (67). In a very recent and thorough examina- icantly associated with training-induced improvements in tion of 7,159 individuals from the Third National Health the insulin-sensitivity index. and Nutrition Examination Survey, several CRP SNPs were Not all such studies revealed a gene–disease association. found to be associated with CRP levels, but only the Thus, although baseline CRP is a good predictor of reste- functional Ϫ390C/T/A SNP was linked to both CRP levels nosis (59), investigation of the ϩ1059G/C and ϩ29T/A and to increased prevalence of coronary heart disease (68). SNPs found no association with restenosis risk after angio- One interpretation of the findings of these studies, which plasty (60). These data consistently point toward an asso- were performed independently using diverse and unrelated ciation of CRP polymorphisms with blood CRP levels. The populations, is that there are some CRP sequence polymor- strength and direction of this association appear to depend phisms that influence expression of the protein that might on the type of disease, however, and the polymorphisms also affect cardiovascular well being. may or may not associate with disease risk. All of the known CRP gene SNPs are in high-linkage The Big Question: disequilibrium, so tag SNPs have been used to define Does CRP Gene Polymorphism Alter extended CRP haplotypes. Using this approach, Kardys CRP Blood Levels and Thus Disease Status? et al. (69) found that although CRP blood level was associated with incident coronary heart disease and varied Zee et al. (61) examined the relationship of ϩ1059G/C with CRP haplotype, the different CRP haplotypes them- with blood CRP and risk of arterial thrombosis. In a large selves were not associated with coronary heart disease. In cohort of healthy American men (726 case-control pairs), another tag SNP investigation of subjects from the NHLBI 12% were found to carry the C allele, and those turned out Family Heart Study, a trend toward a significant association to have lower median CRP values compared with indivi- of CRP haplotypes with blood CRP was reported. In this duals having the GG genotype. No correlation with arterial case, the Ϫ757T/C and Ϫ390C/T/A SNPs showed the thrombosis was found (61). Likewise, in the British strongest association. Again, however, none of the CRP Women’s Heart and Health Study (62), carriers of ϩ1059C haplotypes or tag SNPs showed an association with intima- had lower blood CRP, but this did not affect blood pressure, media thickness as measured by ultrasound (70). Another pulse pressure, or hypertension. In contrast, Balistreri et al. study examining the correlation of various CRP SNPs to (63) found that CRP blood level was higher in subjects with CRP blood level and arterial pulse wave velocity (71) found 1120 Hage and Szalai JACC Vol. 50, No. 12, 2007 CRP Gene Polymorphisms September 18, 2007:1115–22 statistically significant association between CRP levels and direct relationship between a particular SNP or CRP Ϫ390C/T/A and ϩ1444C/T, and there was association haplotype and actual clinical events. of ϩ1059G/C with pulse-wave velocity. Finally, in the Coronary artery disease is considered by many to be a Cardiovascular Health Study, ϩ1846C/T, ϩ29A/T, disease of inflammation (73), and blood CRP has been ϩ1059G/C, and Ϫ757T/C were variably associated with shown to be a powerful predictor of cardiovascular events blood CRP. Not surprisingly, given that the distribution of (74,75). The effects of lifestyle and trans-acting factors not CRP alleles differed among the races, the disease risk also withstanding, it is now increasingly apparent that certain differed among the races. Thus, in Caucasians ϩ29A/T was subtle variations in the CRP gene sequence predictably and associated with increased risk of stroke and cardiovascular strongly influence blood CRP level. Furthermore, some of mortality and ϩ1059G/C and ϩ1846C/T were associated these variations map to cis-acting elements that have already with cardiovascular mortality, whereas in African Ameri- been associated with clinical correlates of cardiovascular cans Ϫ757T/C was associated with myocardial infarctions disease; notable in this respect is the triallelic SNP at (72). At least some of the large-scale investigations using tag position Ϫ390. New CRP SNPs and novel gene-protein- SNP approaches have provided some evidence that se- disease associations will likely be reported at an increasing quence differences in the CRP gene alter both CRP blood pace, so the exact contribution of CRP gene polymorphisms levels and disease risk. to the protein’s level and to disease status will continue to define itself. However, reports that individual SNPs in CRP Summary might explain up to 10% of the variance in blood CRP level, and that the heritability of blood CRP is 30% to 40%, We have attempted here to bring together for comparison suggest that identifying an individual’s CRP SNPs might and contrast all of the published reports showing (or not one day be a practical and useful clinical approach for showing) that CRP gene polymorphisms associate with prediction of cardiovascular event risk or for selecting CRP blood levels or/and that CRP gene polymorphisms patients for treatment with CRP targeting drugs. Of course, associate with cardiovascular disease risk. Some noteworthy moving CRP genotyping to the clinical arena depends patterns emerged. There is now ample direct evidence that entirely on obtaining proof that CRP participates in disease some CRP gene polymorphisms affect the amount of CRP processes in humans, and that proof is still not available. If produced and indirect evidence that this in turn might have ultimately CRP is found to not contribute to cardiovascular an impact on vascular disease. In the CRP gene, intron biology, its utility as a marker of inflammation will not be variations in the GT repeat have been linked to changes in diminished. Ongoing and future studies will no doubt settle blood CRP and the risk of vascular events, and the ϩ29A/T the issue, but until then, there will always be critics and SNP was linked to blood CRP level but not to restenosis. In supporters of the position that CRP blood levels and CRP exon II, the often-studied ϩ1059G/C SNP has been shown gene polymorphisms really matter. to correlate well with blood CRP and in at least 1 study to associate with pulse-wave velocity. It does not correlate, Reprint requests and correspondence: Dr. Fadi G. Hage, 306 however, with arterial thrombosis, blood pressure, or reste- Lyons-Harrison Research Building, 1530 3rd Avenue South, nosis. In the promoter region, the Ϫ390C/T/A triallelic Birmingham, Alabama 35294-0007. E-mail: [email protected]. SNP, the only functional CRP polymorphism so far de- scribed, has been shown to correlate with blood CRP and recently has been directly linked to risk of cardiovascular REFERENCES disease. The other SNPs in the promoter region have 1. Tillett W, Francis T. Serological reactions in pneumonia with non- variable association with CRP level but good correlation protein somatic fraction of pneumococcus. J Exp Med 1930;52: with cardiac disease, diabetes mellitus, and insulin resis- 561–71. 2. Pepys MB, Baltz ML. Acute phase proteins with special reference to tance. In the 3= untranslated and 3= flanking regions there C-reactive protein and related proteins (pentaxins) and serum amyloid are other SNPs that also correlate with blood CRP, but A protein. Adv Immunol 1983;34:141–212. none that associate with cardiovascular disease (although 3. Volanakis JE. Human C-reactive protein: expression, structure, and ϩ function. Mol Immunol 2001;38:189–97. 1846C/T does associate with SLE). Importantly, nearly 4. van Leeuwen MA, van Rijswijk MH. Acute phase proteins in the 10% of the total variation in CRP blood level can be monitoring of inflammatory disorders. Baillieres Clin Rheumatol accounted for by the CRP gene SNPs. This is remarkable 1994;8:531–52. 5. Bassuk SS, Rifai N, Ridker PM. High-sensitivity C-reactive protein: given that CRP blood level is as much as 40% heritable. The clinical importance. Curr Probl Cardiol 2004;29:439–93. fact that most studies have not been able to link CRP 6. Pepys MB, Hirschfield GM, Tennent GA, et al. Targeting C-reactive polymorphisms to events is not that surprising, despite their protein for the treatment of cardiovascular disease. Nature 2006;440: 1217–21. clear relationship to blood CRP level and of the latter to 7. Wang D, Oparil S, Chen YF, et al. Estrogen treatment abrogates hard clinical end points. Given the modest proportion of neointima formation in human C-reactive protein transgenic mice. variation in blood CRP level that CRP polymorphisms Arterioscler Thromb Vasc Biol 2005;25:2094–9. 8. Danenberg HD, Szalai AJ, Swaminathan RV, et al. Increased throm- explain, it would take a very large study (several orders of bosis after arterial injury in human C-reactive protein-transgenic mice. magnitude larger than those so far undertaken) to see a Circulation 2003;108:512–5. JACC Vol. 50, No. 12, 2007 Hage and Szalai 1121 September 18, 2007:1115–22 CRP Gene Polymorphisms

9. Griselli M, Herbert J, Hutchinson WL, et al. C-reactive protein and death after an acute coronary syndrome in male elderly patients. Int complement are important mediators of tissue damage in acute J Cardiol 2005;103:266–71. myocardial infarction. J Exp Med 1999;190:1733–40. 33. Szalai AJ, van Ginkel FW, Dalrymple SA, Murray R, McGhee JR, 10. Kushner I, Rzewnicki D, Samols D. What does minor elevation of Volanakis JE. and IL-6 requirements for human C-reactive protein signify? Am J Med 2006;119:166, e17–28. C-reactive protein gene expression in transgenic mice. J Immunol 11. Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive 1998;160:5294–9. protein, interleukin 6, and risk of developing type 2 diabetes mellitus. 34. Weinhold B, Bader A, Poli V, Ruther U. Interleukin-6 is necessary, JAMA 2001;286:327–34. but not sufficient, for induction of the human C-reactive protein gene 12. Laaksonen DE, Niskanen L, Nyyssonen K, et al. C-reactive protein in vivo. Biochem J 1997;325:617–21. and the development of the metabolic syndrome and diabetes in 35. Lakoski SG, Cushman M, Criqui M, et al. Gender and C-reactive middle-aged men. Diabetologia 2004;47:1403–10. protein: data from the Multiethnic Study of Atherosclerosis (MESA) 13. Haffner SM. The metabolic syndrome: inflammation, diabetes melli- cohort. Am Heart J 2006;152:593–8. tus, and cardiovascular disease. Am J Cardiol 2006;97:3A–11A. 36. Venugopal SK, Devaraj S, Jialal I. Macrophage conditioned medium 14. Esposito K, Pontillo A, Di Palo C, et al. Effect of weight loss and induces the expression of C-reactive protein in human aortic endo- lifestyle changes on vascular inflammatory markers in obese women: a thelial cells: potential for paracrine/autocrine effects. Am J Pathol randomized trial. JAMA 2003;289:1799–804. 2005;166:1265–71. 15. Kluft C, de Maat MP. Genetics of C-reactive protein: new possibilities 37. Pankow JS, Folsom AR, Cushman M, et al. Familial and genetic and complications. Arterioscler Thromb Vasc Biol 2003;23:1956–9. determinants of systemic markers of inflammation: the NHLBI family 16. Kathiresan S, Larson MG, Vasan RS, et al. Contribution of clinical heart study. Atherosclerosis 2001;154:681–9. correlates and 13 C-reactive protein gene polymorphisms to interin- 38. Worns MA, Victor A, Galle PR, Hohler T. Genetic and environmen- dividual variability in serum C-reactive protein level. Circulation tal contributions to plasma C-reactive protein and interleukin-6 2006;113:1415–23. levels—a study in twins. Genes Immunol 2006;7:600–5. 17. Hingorani AD, Shah T, Casas JP. Linking observational and genetic 39. MacGregor AJ, Gallimore JR, Spector TD, Pepys MB. Genetic effects approaches to determine the role of C-reactive protein in heart disease on baseline values of C-reactive protein and serum amyloid a protein: risk. Eur Heart J 2006;27:1261–3. a comparison of monozygotic and dizygotic twins. Clin Chem 2004; 18. Whitehead AS, Bruns GA, Markham AF, Colten HR, Woods DE. 50:130–4. Isolation of human C-reactive protein complementary DNA and 40. Retterstol L, Eikvar L, Berg K. A twin study of C-reactive protein localization of the gene to chromosome 1. Science 1983;221:69–71. compared to other risk factors for coronary heart disease. Atheroscle- 19. Floyd-Smith G, Whitehead AS, Colten HR, Francke U. The human rosis 2003;169:279–82. C-reactive protein gene (CRP) and serum amyloid P component gene 41. Weber JL, Kwitek AE, May PE. Dinucleotide repeat polymorphism at (APCS) are located on the proximal long arm of chromosome 1. the CRP locus. Nucleic Acids Res 1990;18:4635. Immunogenetics 1986;24:171–6. 42. Cao H, Hegele RA. Human C-reactive protein (CRP) 1059G/C 20. Walsh MT, Divane A, Whitehead AS. Fine mapping of the human polymorphism. J Hum Genet 2000;45:100–1. pentraxin gene region on chromosome 1q23. Immunogenetics 1996; 43. Hegele RA, Ban MR, Young TK. Serum C-reactive protein in 44:62–9. Canadian Inuit and its association with genetic variation on chromo- 21. Woo P, Korenberg JR, Whitehead AS. Characterization of genomic some 1q21. Clin Chem 2001;47:1707–9. and complementary DNA sequence of human C-reactive protein, and 44. Crawford DC, Yi Q, Smith JD, et al. Allelic spectrum of the natural comparison with the complementary DNA sequence of serum amyloid variation in CRP. Hum Genet 2006;119:496–504. P component. J Biol Chem 1985;260:13384–8. 45. Szalai AJ, McCrory MA, Cooper GS, Wu J, Kimberly RP. Associa- 22. Lei KJ, Liu T, Zon G, Soravia E, Liu TY, Goldman ND. Genomic tion between baseline levels of C-reactive protein (CRP) and a DNA sequence for human C-reactive protein. J Biol Chem 1985;260: dinucleotide repeat polymorphism in the intron of the CRP gene. 13377–83. Genes Immunol 2002;3:14–9. 23. Goldman ND, Liu T, Lei KJ. Structural analysis of the locus 46. Brull DJ, Serrano N, Zito F, et al. Human CRP gene polymorphism containing the human C-reactive protein gene and its related pseudo- influences CRP levels: implications for the prediction and pathogen- gene. J Biol Chem 1987;262:7001–5. esis of coronary heart disease. Arterioscler Thromb Vasc Biol 2003; 24. Szalai AJ, Agrawal A, Greenhough TJ, Volanakis JE. C-reactive 23:2063–9. protein: structural biology, gene expression, and host defense function. 47. D’Aiuto F, Casas JP, Shah T, Humphries SE, Hingorani AD, Tonetti Immunol Res 1997;16:127–36. MS. C-reactive protein (ϩ1444CϾT) polymorphism influences CRP 25. Toniatti C, Arcone R, Majello B, Ganter U, Arpaia G, Ciliberto G. response following a moderate inflammatory stimulus. Atherosclerosis Regulation of the human C-reactive protein gene, a major marker of 2005;179:413–7. inflammation and cancer. Mol Biol Med 1990;7:199–212. 48. Marsik C, Sunder-Plassmann R, Jilma B, et al. The C-reactive protein 26. Kushner I, Jiang SL, Zhang D, Lozanski G, Samols D. Do post- ϩ1444C/T alteration modulates the inflammation and transcriptional mechanisms participate in induction of C-reactive response in human endotoxemia. Clin Chem 2006;52:1952–7. protein and serum amyloid A by IL-6 and IL-1? Ann N Y Acad Sci 49. Obisesan TO, Leeuwenburgh C, Phillips T, et al. C-reactive protein 1995;762:102–7. genotypes affect baseline, but not exercise training-induced changes, in 27. Berger P, McConnell JP, Nunn M, et al. C-reactive protein levels are C-reactive protein levels. Arterioscler Thromb Vasc Biol 2004;24: influenced by common IL-1 gene variations. Cytokine 2002;17:171–4. 1874–9. 28. Latkovskis G, Licis N, Kalnins U. C-reactive protein levels and 50. Suk HJ, Ridker PM, Cook NR, Zee RY. Relation of polymorphism common polymorphisms of the interleukin-1 gene cluster and within the C-reactive protein gene and plasma CRP levels. Athero- interleukin-6 gene in patients with coronary heart disease. Eur sclerosis 2005;178:139–45. J Immunogenet 2004;31:207–13. 51. Szalai AJ, Wu J, Lange EM, et al. Single-nucleotide polymorphisms in 29. Vickers MA, Green FR, Terry C, et al. Genotype at a promoter the C-reactive protein (CRP) gene promoter that affect transcription polymorphism of the interleukin-6 gene is associated with baseline factor binding, alter transcriptional activity, and associate with differ- levels of plasma C-reactive protein. Cardiovasc Res 2002;53:1029–34. ences in baseline serum CRP level. J Mol Med 2005;83:440–7. 30. Chiappelli M, Tampieri C, Tumini E, et al. Interleukin-6 gene 52. Kovacs A, Green F, Hansson LO, et al. A novel common single polymorphism is an age-dependent risk factor for myocardial infarc- nucleotide polymorphism in the promoter region of the C-reactive tion in men. Int J Immunogenet 2005;32:349–53. protein gene associated with the plasma concentration of C-reactive 31. Iacoviello L, Di Castelnuovo A, Gattone M, et al. Polymorphisms of protein. Atherosclerosis 2005;178:193–8. the interleukin-1beta gene affect the risk of myocardial infarction and 53. Morrow DA, Rifai N, Antman EM, et al. C-reactive protein is a ischemic stroke at young age and the response of mononuclear cells to potent predictor of mortality independently of and in combination stimulation in vitro. Arterioscler Thromb Vasc Biol 2005;25:222–7. with troponin T in acute coronary syndromes: a TIMI 11A substudy. 32. Antonicelli R, Olivieri F, Bonafe M, et al. The interleukin-6 Thrombolysis In Myocardial Infarction. J Am Coll Cardiol 1998;31: Ϫ174GϾC promoter polymorphism is associated with a higher risk of 1460–5. 1122 Hage and Szalai JACC Vol. 50, No. 12, 2007 CRP Gene Polymorphisms September 18, 2007:1115–22

54. James SK, Armstrong P, Barnathan E, et al. Troponin and C-reactive 65. Russell AI, Cunninghame Graham DS, Shepherd C, et al. Polymor- protein have different relations to subsequent mortality and myocardial phism at the C-reactive protein locus influences gene expression and infarction after acute coronary syndrome: a GUSTO-IV substudy. predisposes to systemic lupus erythematosus. Hum Mol Genet 2004; J Am Coll Cardiol 2003;41:916–24. 13:137–47. 55. Danik JS, Chasman DI, Cannon CP, et al. Influence of genetic 66. Chen J, Zhao J, Huang J, Su S, Qiang B, Gu D. Ϫ717AϾG variation in the C-reactive protein gene on the inflammatory response polymorphism of human C-reactive protein gene associated with during and after acute coronary ischemia. Ann Hum Genet 2006;70: coronary heart disease in ethnic Han Chinese: the Beijing atheroscle- 1–12. rosis study. J Mol Med 2005;83:72–8. 56. Roy S, Hill AV, Knox K, Griffiths D, Crook D. Research pointers: 67. Miller DT, Zee RY, Suk Danik J, et al. Association of common CRP association of common genetic variant with susceptibility to invasive gene variants with CRP levels and cardiovascular events. Ann Hum pneumococcal disease. BMJ 2002;324:1369. Genet 2005;69:623–38. 57. Wolford JK, Gruber JD, Ossowski VM, et al. A C-reactive protein 68. Crawford DC, Sanders CL, Qin X, et al. Genetic variation is promoter polymorphism is associated with type 2 diabetes mellitus in associated with C-reactive protein levels in the third National Health Pima Indians. Mol Genet Metab 2003;78:136–44. and Nutrition Examination Survey. Circulation 2006;114:2458–65. 58. Obisesan TO, Leeuwenburgh C, Ferrell RE, et al. C-reactive protein 69. Kardys I, de Maat MP, Uitterlinden AG, Hofman A, Witteman JC. genotype affects exercise training-induced changes in insulin sensitiv- C-reactive protein gene haplotypes and risk of coronary heart disease: ity. Metabolism 2006;55:453–60. the Rotterdam Study. Eur Heart J 2006;27:1331–7. 59. Buffon A, Liuzzo G, Biasucci LM, et al. Preprocedural serum levels of 70. Wang Q, Hunt SC, Xu Q, et al. Association study of CRP gene C-reactive protein predict early complications and late restenosis after polymorphisms with serum CRP level and cardiovascular risk in the coronary angioplasty. J Am Coll Cardiol 1999;34:1512–21. NHLBI Family Heart Study. Am J Physiol Heart Circ Physiol 60. Zee RY, Hegener HH, Fernandez-Cruz A, Lindpaintner K. C-reactive protein gene polymorphisms and the incidence of post- 2006;291:H2752–7. angioplasty restenosis. Atherosclerosis 2004;176:393–6. 71. Morita A, Nakayama T, Doba N, Hinohara S, Soma M. Polymor- 61. Zee RY, Ridker PM. Polymorphism in the human C-reactive protein phism of the C-reactive protein (CRP) gene is related to serum CRP (CRP) gene, plasma concentrations of CRP, and the risk of future Level and arterial pulse wave velocity in healthy elderly Japanese. arterial thrombosis. Atherosclerosis 2002;162:217–9. Hypertens Res 2006;29:323–31. 62. Davey Smith G, Lawlor DA, Harbord R, et al. Association of 72. Lange LA, Carlson CS, Hindorff LA, et al. Association of polymor- C-reactive protein with blood pressure and hypertension: life course phisms in the CRP gene with circulating C-reactive protein levels and confounding and mendelian randomization tests of causality. Arterio- cardiovascular events. JAMA 2006;296:2703–11. scler Thromb Vasc Biol 2005;25:1051–6. 73. Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 63. Balistreri CR, Vasto S, Listi F, et al. Association between ϩ1059G/C 1999;340:115–26. CRP polymorphism and acute myocardial infarction in a cohort of 74. Danesh J, Wheeler JG, Hirschfield GM, et al. C-reactive protein and patients from Sicily: a pilot study. Ann N Y Acad Sci 2006;1067: other circulating markers of inflammation in the prediction of coronary 276–81. heart disease. N Engl J Med 2004;350:1387–97. 64. Szalai AJ, Alarcon GS, Calvo-Alen J, et al. Systemic lupus erythem- 75. Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of atosus in a multiethnic U.S. cohort (LUMINA). XXX: association C-reactive protein and low-density lipoprotein cholesterol levels in the between C-reactive protein (CRP) gene polymorphisms and vascular prediction of first cardiovascular events. N Engl J Med 2002;347: events. Rheumatol (Oxf) 2005;44:864–8. 1557–65.