Watala C. Regulation of cell function by isoforms of C-reactive protein: a comparative analysis

Despite the emerging evidence suggesting a proatherogenic role of C-reactive protein (CRP) in atherosclerosis, the contribution of CRP in pathogenesis of atherosclerosis and atherothrombosis has not been unequivocally defined. The role of CRP in pathophysiology/pathology seems to largely depend on its structure. Two CRP isoforms, the native pentameric and the modified monomeric one, differ substantially in their physiological functions, which is thought to origin from the considerable structural heterogeneity of the CRP molecule. The present review provides an overview of the experimental evidence with relevance to the clinical role(s) of various CRP isoforms. The biological role of the protein, its structure and distribution are discussed with particular emphasis on the diverse properties of the pentameric and monomeric forms of CRP. Some methodological aspects, related to experimental models and techniques of CRP preparation, are also critically reviewed.


Biology of C-ReACTive PRoTein
C-reactive protein was originally discovered by Tillett and Francis (1930) in sera from patients with Streptococcus pneumonia infection.Its name derives from the interaction with phosphocholine (PCh) residues of pneumococcal C-polysaccharide (PnC), a component of teichoic acid in the pneumococcal cell wall.The binding of phosphocholine and the complement pathway component (C1q) by CRP is part of innate immunity that activates the classical com-plement pathway (Gabay & Kushner, 1999;Du Clos, 2000).The interactions between CRP and its various ligands, like phosphocholine or Fcγ receptors, seem to be multifaceted processes that, apart from C1q, may affect microorganisms, apoptotic cells (Gershov et al., 2000), damaged cell membranes (Volanakis & Wirtz, 1979), phagocytic cells (Ballou & Lozanski, 1992), smooth muscle cells (Hattori et al., 2003), endothelial progenitor cells (Verma et al., 2004a) and endothelial cells (Pasceri et al., 2000), thereby modulating the inflammatory response.

2009
M. Boncler and C. Watała STRuCTuRe of The moleCule C-Reactive protein is a member of the phylogenetically highly conserved family of proteinspentraxins.The name pentraxin was first assigned to CRP due to the recognition of its ultrastructural appearance and the radial symmetry of five CRP subunits forming a pentagon ring.In general, pentraxins are characterized by a cyclic multimeric structure, calcium-dependent ligand binding, and a distinctive flattened β-jellyroll structure similar to that of the legume lectins (Emsley et al., 1994).Together with serum amyloid P component (SAP), C-reactive protein belongs to the group of short pentraxins, which differ from long pentraxins mainly by the primary structure of the subunits, gene organization, cellular source, and ligand-binding properties (Garlanda et al., 2005).Human CRP gene is localized on chromosome 1 and consists of two exons (Whitehead et al., 1983).

ACuTe PhASe ReSPonSe
A native CRP molecule comprises five identical noncovalently bound subunits of 206 aa and a molecular mass of about 23 kDa each (Osmand et al., 1977).All subunits have the same orientation in the pentamer, with a PCh binding site located on the so-called "recognition face" of each subunit.The active PCh-binding site consists of a hydrophobic pocket formed by the residues of Leu 64 , Phe 66 , and Thr 76 , two calcium ions, and Glu 81 , which is located on the other side of the hydrophobic pocket (Shrive et al., 1996).Crystallographic data on CRP-PCh complexes, together with the findings on the mutational analysis of the PCh-binding site reported by Agrawal et al. (2002), clearly point out to crucial importance of Phe 66 and Glu 81 in mediating the binding of phosphocholine to CRP.Whereas the former provides hydrophobic interactions with methyl groups of choline, the latter interacts with the positively charged choline nitrogen.CRP complexed with C-polysaccharide, phospholipid ligands, or protamine is recognized by the complement pathway component (C1q) at the opposite side of the CRP pentamer ("effector face") -and thus efficiently activates the classical complement pathway.Studies on a series of CRP mutants have identified several residues that are critical for the binding of CRP to C1q (Agrawal & Volanakis, 1994).Among them, Asp 112 at the open end of the cleft, where a well-defined shallow pocket is formed -is regarded as a major determinant of C1q binding to CRP complexed with PCh.Other, equally important contact sites seem to be Lys 114 and Tyr 175 .Two mutant forms of CRP, K114A (sub-stitution of Ala for Lys 114 ) and Y175A (substitution of Ala for Tyr 175 ) bind PnC as avidly as the wildtype CRP but they differ in their binding to C1q.Whereas the Y175A form does not interact with human C1q, K114A binds to C1q and activates human complement many-fold better than the wild-type CRP (Agrawal & Volanakis, 1994;Agrawal et al., 2001).It has also been shown that neither the wildtype CRP nor the CRP mutants (K114A or Y175A) interact with mouse C1q (Suresh et al., 2006).The site(s) of C1q that is responsible for the interaction with CRP is poorly defined and is referred to different regions of C1q.This molecule can bind various activators, including IgG and/or IgM, CRP, and SAP; however, the C1q-binding sites for immunoglobulins have been mapped and attributed to the globular head region (C1qGR) (Duncan & Winter, 1988;Perkins et al., 1991), whereas those for some other ligands were shown to be located within its collagen-like region (C1qCLR).Initially, the latter was reported to be responsible for the interaction with CRP (Jiang et al., 1991).Nevertheless, recent competition studies with the use of several mAbs directed against C1qGR or C1qCLR of human C1q have revealed that it is the globular head region of C1q that is involved in the C1q-CRP interaction (McGrath et al., 2006).

C-ReACTive PRoTein ReCePToRS
CRP exhibits multiple functional similarities to the G class of immunoglobulins (IgG), including their ability to interact with both C1q and Fcγ receptors (Du Clos, 2000).Whether CRP binds to these ligands at the IgG ligand-binding site is not clear.It is known that IgG subclasses with the highest affinity for the receptor have the sequence 234 LLGGP, which is important for FcγRI (CD64) and FcγRIIa (CD32) binding.A similar sequence 175 YLGGP has been found in CRP molecule, and this site has been proposed to be involved in binding to FcγRI.The modification of Leu 176 to Glu in CRP ( 175 YEGGP-CRP) was found not to decrease CRP binding to monocytes; however, there was no binding of 175 YEGGP-CRP to cells transfected with the cDNA for FcγRI (Marnell et al., 1995).In addition to Leu 176 , the residues Lys 114 , Thr 173 , and Asn 186 in CRP have been indicated to be critical for binding to FcγRI.Thr 173 and Asn 186 were important for binding to FcγRIIa, and mutations of Lys 114 , Leu 176 , and Thr 173 affected C1q binding as well (Bang et al., 2005).Such analysis of CRP mutants identified new important residues in FcγR and C1q binding and provided the evidence that the binding sites on CRP for FcγRI, FcγRIIa, and C1q may overlap.Other intrinsic CRP ligands include small ribonucleoprotein particles (Du Clos, Distinct effects of CRP isoforms on cell function 1989), lipoproteins (Bhakdi et al., 1999;Chang et al., 2002), and extracellular matrix components (Salonen et al., 1984;Tseng & Mortensen, 1988;Swanson et al., 1989).

iSofoRmS of C-ReACTive PRoTein
Native pentameric C-reactive protein (nCRP) may undergo structural changes from pentamer to forms resembling more the free CRP subunit, designated as modified CRP (mCRP).They are distinguished from nCRP by their antigenic, electrophoretic, and biological activities (see below) (Potempa et al., 1983;1987).Generally, mCRP can be generated from nCRP by exposing nCRP to heat, urea, or acidic conditions in the absence of calcium ions.Complete conversion of nCRP to mCRP under denaturating conditions proceeds rapidly (2 min), but it requires high concentrations of urea (Potempa et al., 1983;Kresl et al., 1998).Moreover, direct immobilization of nCRP on polystyrene plates has been shown to produce mCRP (Potempa et al., 1983).On membranes, including those of liposomes and cells, such a structural transition of nCRP into mCRP may be a more complicated process that involves the formation of monomers that express mCRP antigenicity but retain the native pentameric conformation (originally designated by the authors as mCRP m ) (Ji et al., 2007).According to these authors, expression of the neoepitope in nCRP bound to a membrane is not necessarily associated with a physical separation of the subunits.The dissociation process may consist of the following steps: (1) nCRP binds to a membrane; (2) nCRP undergoes conformational changes on the membrane such as slight but appreciable tertiary structure alteration and neoepitope expression; (3) relaxation of nCRP oligomeric structure; and (4) physical separation of subunits.The extent of nCRP dissociation strongly depends on CRP ligand and experimental conditions.As suggested, a hydrophobic microenvironment, considerable ligand mobility (relevant to bilayer lipid fluidity), and multipoint attachment are necessary elements for efficient dissociation of nCRP.In the monomeric form, CRP retains the ability to interact with complement (Ji et al., 2006b;Biro et al., 2007).Depending on whether mCRP is in a fluid phase or is bound to a surface, there are two pathways by which mCRP may bind to C1q and thus regulate complement activation.When mCRP is in the ligand-free state, it exhibits an inhibitory activity towards a complement, whereas immobilized mCRP is able to activate the classical complement pathway, thus demonstrating a dual role in the innate immune system (Ji et al., 2006b).

C-ReACTive PRoTein SynTheSiS
Native CRP is synthesized in a soluble form by hepatocytes from where it is secreted into the circulation (Hurlimann et al., 1966).Interleukin 1 (IL-1), IL-6, and tumor necrosis factor (TNF) are considered important mediators for the modulation of CRP synthesis in the liver (Ganter et al., 1989;Yap et al., 1991).Induction of CRP expression in human hepatoma Hep3B cells by proinflammatory cytokines (IL-6 and IL-1β) may be hampered by statins or nitric oxide (NO) (Voleti & Agrawal, 2006).Furthermore, CRP has been found in extrahepatic tissues; however, it remains unknown whether it is deposited at those sites from a circulating plasma pool or is produced locally by the cells, and what exactly is the contribution of extrahepatic CRP to its serum levels.Arterial tissue itself is able to produce CRP and complement proteins, and both the mRNAs and proteins are substantially up-regulated in atherosclerotic plaques (Yasojima et al., 2001).The first evidence for the presence of CRP in atherosclerotic lesion was revealed in immunohistochemical studies (Vlaicu et al., 1985), and experimentation along this line brings new insights into the nature of CRP.Significant levels of human CRP antigen were found in human early atherosclerotic lesions with the location in deep fibroelastic layer and fibromuscular layer of the intima adjacent to the media (Torzewski et al., 1998).At these sites, CRP is frequently colocalized with the terminal C5b-9 component complex, which supports the hypothesis of in situ CRP-mediated complement activation in the arterial wall (Torzewski et al., 1998).Unexpectedly, that conclusion became a subject of controversy since the authors had used in their study a monoclonal antibody that recognized both isoforms of CRP, the native and the modified one.Following the authors' suggestions, it is likely that nCRP may activate complement after binding to ligands (phospholipids, lipoproteins, or nuclear debris), and it may finally be transformed into mCRP under the acidic conditions of an inflammatory microenvironment.One cannot exclude that such a modification of the protein may occur at least in the course of intermediate and advanced atherosclerotic lesions (Vaith & Potempa, 2000).
Modified CRP seems to be a naturally occurring tissue-based form of CRP in the organism, and it has been identified in both normal and pathological tissues.By probing various human tissues (pulmonary, ovarian, testicular, cardiac) with well characterized monoclonal antibodies specific for either nCRP or mCRP, it has been demonstrated that significant amounts of mCRP are present in the walls of blood vessels associated with normal human tissues (Diehl et al., 2000).However, normal renal tissue was found to be mCRP-negative.In turn, diabetic patients with severe chronic kidney disease showed progressive tubular mCRP staining with declining renal function and increasing severity of histological lesions (Schwedler et al., 2003).The natural expression of CRP in the cells involved in atherosclerotic lesion formation has also been investigated (Ciubotaru et al., 2005).CRP transcript was minimally expressed in undifferentiated macrophages (U937-derived macrophages).The expression of CRP increased markedly in macrophages during differentiation from monocytes, and it was not affected by LPS at 24 h.Further investigation of the effect of LPS on the production of CRP protein showed that LPS time-dependently increased the production of CRP, peaking at 48 h.After 24 h of LPS stimulation, macrophages had a significantly higher amount of CRP compared with unstimulated, differentiated macrophages.These findings indicate that LPS modulates CRP synthesis at the translational rather than transcriptional level, since there was no concurrent increase in the level of CRP mRNA detected in PMA-differentiated macrophages in the presence or absence of LPS stimulation.In addition, LPS-stimulated macrophages most likely produce mCRP.Other investigators reported the presence of CRP in human neuronal cells (Yasojima et al., 2000), lung epithelial cells (Ramage et al., 2004), renal cortical tubular epithelial cells (Jabs et al., 2003), and human coronary artery smooth muscle cells (Calabro et al., 2003).

C-ReACTive PRoTein And CARdiovASCulAR diSeASe
C-reactive protein is a classical acute-phase plasma protein, and its serum levels rise rapidly and markedly (even to as much as 1000-fold within 48 h) in response to bacterial infection, trauma, tissue necrosis, and inflammation (Pepys, 1981;Gewurz et al., 1982).In recent years, CRP, formerly considered solely as a marker of inflammation, has been designated as a powerful independent predictor of future cardiovascular risk.A plethora of studies have demonstrated a direct association between an elevated CRP level and the incidence of cardiovascular complications in individuals without overt cardiovascular disease (Ridker et al., 2000a;2000b), as well as in patients with unstable angina (Liuzzo et al., 1994;Biasucci et al., 1999), myocardial infarction (Ridker et al., 2003), ischemic stroke (Ridker, 2002a), or peripheral artery disease (Ridker et al., 2001).In addition, an elevated CRP concentration in blood serum predicts the risk of sudden death (Albert et al., 2002) and restenosis after percutaneous coronary intervention (Blake & Ridker, 2002).The outcomes of the Women's Health Study suggest that determina-tion of CRP in blood may provide a valuable tool in primary prevention as it permits identifying apparently healthy subjects who are at risk of developing cardiovascular events in conjunction with an altered lipid profile (Ridker et al., 2002b).Furthermore, CRP seems to be an even stronger predictor of the risk of cardiovascular events than serum levels of LDL cholesterol.Despite the emerging evidence suggesting a proatherogenic role of CRP in atherosclerosis, the contribution of CRP in pathogenesis of atherosclerosis and atherothrombosis is not yet unequivocally defined and accepted (Nilsson, 2005;Pepys, 2005;2008;Scirica & Morrow, 2006).Very recently, the relationship between CRP level, several CRP polymorphisms and risk of ischaemic vascular disease has been analyzed in four large independent studies (Zacho et al., 2008).Consistent with most but not all previous studies are reports pointing out that the genetic variants associated with life-long increases in plasma CRP levels were not associated with an increased risk of ischaemic vascular disease (Kovacs et al., 2005;Wang et al., 2006;Kathiresan et al., 2006;Reitz et al., 2007;Bis et al., 2008).In the light of a concurrent observation on a significant association between CRP level and an increased risk of ischaemic heart disease and ischaemic cerebrovascular disease, such findings suggest that increased CRP level is not a cause but rather a marker for atherosclerosis (Zacho et al., 2008).

In vItro STudieS
Numerous literature reports document the role of C-reactive protein in atherogenesis.Epidemiological evidence reveals an association between elevated CRP plasma levels and atherosclerosis (Haverkate et al., 1997;Ridker et al., 1997;Koenig et al., 1999).Infusion of recombinant CRP in healthy men results in the activation of inflammation and coagulation (Bisoendial et al., 2005).In vitro, CRP has been shown to exert direct proinflammatory and proatherosclerotic effects on vascular cells, as exemplified by: (1) induction of an increased expression of adhesion molecules, such as vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1), and E-selectin (Pasceri et al., 2000); (2) stimulation of secretion of monocyte chemoattractant protein-1 (MCP-1) (Pasceri et al., 2001); and (3) facilitation of macrophage LDL uptake (Verma et al., 2002a).At present, we think that the majority of these effects are likely to be mediated, in part, by an increased secretion of the potent endothelium-derived vasoactive factor, endothelin-1 (ET-1), and the inflammatory cytokine, interleukin-6 (IL-6), because these effects are attenuated by both bosentan (mixed ET A/B receptor antagonist) and monoclonal anti-hu-Distinct effects of CRP isoforms on cell function man IL-6 antibodies (Verma et al., 2002a).C-Reactive protein-induced endothelial dysfunction may also be mediated by a major receptor for ox-LDL, the endothelial lectin-like oxidized LDL receptor-1 (LOX-1).Incubation of human aortic endothelial cells (HAECs) with CRP leads to enhanced synthesis of LOX-1 mRNA and protein, and the observed changes become attenuated in the presence of anti-ET-1 or IL-6 antibodies, thus supporting the above concept on the association between the stimulated ET-1 and IL-6 production and biological effects of CRP (Li et al., 2004).
The conformation of the CRP molecule seems to be of crucial importance for triggering of the release of key regulators of leukocyte recruitment and expression of adhesion molecules in endothelial cells.Khreiss et al. (2004a) have demonstrated that modified CRP has a proinflammatory effect on human coronary artery endothelial cells (HCAECs) by increasing MCP-1 and IL-8 secretion and by inducing the cellular expressions of ICAM-1, VCAM-1, and Eselectin following a 4-hour incubation with the protein.Otherwise, the contribution of native CRP in the development of vascular inflammation has been suggested to be a minor one: native CRP was not able to promote proinflammatory HCAEC phenotype until the incubation time was prolonged from 4 h to 24 h.In studies on platelet adhesion to bovine aortic endothelial cells (BAECs), the pro-adhesive effect of CRP was also time-dependent, requiring at least 7 h of incubation (Yaron et al., 2006).This remains fairly consistent with previous studies pointing to a pro-adhesive action of nCRP merely after 6-12 h of incubation with endothelium, with the maximal effects at 24 h (Pasceri et al., 2000;2001;Verma et al., 2002a).Interestingly, these findings coincide very well with the outcome of in vitro studies on the kinetics of CRP dissociation into subunits (Wang & Sui, 2001).In another comparative analysis of CRP conformers, the native CRP isoform was even more potent than modified CRP as the modulator of IL-8, PAI-1, cGMP, and prostaglandin F1-α levels after 6 h of incubation with HAECs (Devaraj et al., 2006).It is worth mentioning that the origin or type of endothelial cells might also be critical for the outcome of the study.For instance, Devaraj et al. (2004) failed to observe a stimulatory effect of CRP on MCP-1 release in HAECs, while such an effect was claimed in experiments with human umbilical vein endothelial cells (HUVECs) (Pasceri et al., 2001).
C-Reactive protein also has the ability to modulate directly the production of endotheliumderived vasoactive factors.It is able to profoundly down-regulate the expression and bioactivity of endothelial nitric oxide synthase (eNOS) (Venugopal et al., 2002), NO production (Verma et al., 2002b), and prostacyclin (PGI 2 ) release (Venugopal et al., 2003).The ability of CRP to attenuate NO release is associated with a markedly suppressed in vitro angiogenesis and promotion of endothelial cell apoptosis (Verma et al., 2002b).Further, interactions between C-reactive protein and endothelium may lead to impaired fibrinolysis.Incubation of HAECs with CRP results in a time-and dosedependent increase in the antigen concentration and activity of secreted PAI-1, as well as elevated concentrations of intracellular PAI-1 protein and mRNA (Devaraj et al., 2003).These alterations, however, are mediated by neither ET-1 nor IL-6.Concomitantly, CRP treatment reduces antigen levels and activity of secreted t-PA and downregulates intracellular concentrations of t-PA.Such an inhibition seems to be mediated by the generation of proinflammatory cytokines, like IL-1β and TNFα (Singh et al., 2005).
The direct proatherogenic effects of CRP extend beyond the endothelium and concern also vascular smooth muscle cells (VSMCs), where it activates angiotensin-1 type receptors (Wang et al., 2003).The primary effector molecule of the renninangiotensin system (RAS), angiotensin II (ANG II), has emerged as a critical hormone that affects the function of virtually all organs, including heart, kidney, vasculature, and brain, showing both beneficial and pathological effects.Most of the known physiological effects of angiotensin II are mediated by angiotensin type 1 receptors (AT1Rs).Once ANG II binds to the AT1R, it activates a series of signaling cascades, which in turn regulate the function of VSMCs, endothelial cells, and cardiac fibroblasts, as well as influence their interaction with the extracellular matrix.Convergence of these cascades of events, in addition to abnormalities in the coagulation system, ultimately lead to atherosclerosis and thrombosis with the development of cardiovascular disease (Mehta & Griendling, 2007).The experiments on VSMCs have shown that CRP potently up-regulates AT1R mRNA and protein and increases the number of the binding sites for angiotensin II on the cell surface.Secondly, CRP markedly promoted VSMC migration/proliferation and increased basal reactive oxygen species (ROS) production.Additionally, it potentiated the effects of ANG II on these processes (Wang et al., 2003).In line with those findings, the detrimental CRP-mediated effects have also been observed in bone marrow-derived endothelial progenitor cells (EPCs), which are responsible for vascular regenerative potential and integrity (Suh et al., 2004;Verma et al., 2004a).The number and migratory activity of circulating EPCs have also been shown to inversely correlate with cardiovascular risk (Shantsila et al., 2007).Thus, the ability of CRP to inhibit EPC differentiation and survival may represent an impor-tant mechanism that further links inflammation to cardiovascular disease (Verma et al., 2004a).
During inflammation, the transendothelial migration of leukocytes is controlled by chemotactic and activating signals, together with the sequential interaction between adhesion molecules and their corresponding ligands.C-reactive protein seems to play an important role in these events.To date, experiments with monocytes have shown that CRP induces the production of inflammatory cytokines (IL-1, IL-6, TNFα, IL-8) (Ballou & Lozanski, 1992;Xie et al., 2005), generation of reactive oxygen species (Zeller & Sullivan, 1992), leads to increased expression of tissue factor (Cermak et al., 1993), and affects cell chemotaxis (Whisler et al., 1986;Kew et al., 1990).Very recently, Hanriot et al. (2008), investigating human monocytes exposed to CRP have confirmed the results of earlier studies on CRP-mediated induction of expression of numerous proinflammatory cytokine genes (with the exception of TNFα) and further evidenced increased expression of mRNA for PAI-2, MCP-1, GRO-α, GRO-β, and the chemokine receptors CCR8 and CXCR4.Upon CRP stimulation, monocytic cells have also been shown to up-regulate their vascular endothelial growth factor A (VEGF-A) expression and, to some extent, de novo protein synthesis.Thus, by favoring the activation of endothelial and smooth muscle cells, CRP has been confirmed to promote angiogenesis (Bello et al., 2008).Simultaneously, CRP has been reported to display anti-inflammatory effects in monocytes through down-regulation of α 2 -macroglobulin expression and up-regulation of liver X receptor (LXR) α expression (Hanriot et al., 2008).Such opposing results may be explained by the existence of as yet unidentified pathways activated by CRP in human monocytes and suggest a much more complex biological role of CRP than previously thought.
C-Reactive protein may be an important mediator of leukocyte behavior (activation and adhesion to endothelial cells) in the vasculature through the modulation of monocyte surface antigens.It has been demonstrated that CRP (both isolated from serum and recombinant) can stimulate expression of the monocytic surface integrin CD11b and downregulate that of CD31 antigen (Woollard et al., 2002).In addition, by up-regulating CD11b/CD18 expression and stimulating neutrophil extracellular signalregulated kinase (ERK) activity, the monomeric CRP (but not nCRP) may participate in the promotion of neutrophil adhesion to HCAECs (Zouki et al., 2001).
Up-regulation of adhesion molecules and cytokines by CRP appears to be at the transcriptional level.Transcription of genes encoding the cell adhesion molecules (VCAM-1, ICAM-1, E-selectin) and chemokines is tightly regulated by the transcription factor NF-κB, which has been implicated as a key mediator of atherosclerosis (Brand et al., 1996;Marumo et al., 1997;Thurberg & Collins, 1998;De Martin et al., 2000).This transcription factor is a DNA-binding protein complex that is usually present in the cytosol as an inactive complex.IκB, an associated protein, renders this complex inactive by shielding the nuclear localization signal.Upon IκB phosphorylation and subsequent degradation, the heterodimeric NF-κB complex translocates from the cytoplasm to the nucleus, where it binds to specific DNA sequences in the promotor region of several genes and up-regulates their transcription.Verma et al. (2003) demonstrated for the first time that Creactive protein directly increased the degradation of IκB and subsequently activated the NF-κB signal transduction pathway in saphenous vein endothelial cells.This mechanism has been confirmed by others in studies on the effect of CRP on VCAM-1, IL-8, or MCP-1 expressions.CRP significantly increased IL-8 synthesis in HAECs by the enhancement of NF-κB activity, and such an up-regulation of IL-8 was reversed in the presence of NF-κB inhibitors, like SN-50, parthenolide, or Bay 11 (Devaraj et al., 2004).A significant reduction of IL-8 production by mCRP following the blockade of NF-κB or NO synthesis has also been observed in human neutrophils (Khreiss et al., 2002b).Stimulation of IL-8 production by mCRP in neutrophils was associated with simultaneously increased superoxide production and NO formation leading to enhanced ONOO -formation and consequently activation of NF-κB and activator protein-1 (AP-1).In human mesangial cells, CRP has been found to induce MCP-1 expression via NF-κB activation mediated, at least in part, by intracellular calcium and reactive oxygen species (Chang et al., 2005).In turn, the NF-κB-dependent VCAM-1 expression induced by CRP in BAECs involves PKC, p38MAPK, and tyrosine kinase (Kawanami et al., 2006).Altogether, CRP appears to exert a wide spectrum of activities in vascular tissues that may promote progression of atherosclerosis.
The majority of reports describe the function of pentameric CRP, but only a few of them differentiate the possible effects of native CRP from those of modified CRP.It has been shown that C-reactive protein may interact with various lipoproteins (native, oxidized, or enzymatically modified) primarily when CRP is in a modified form and not the pentameric structure (Ji et al., 2006a).The mechanism of CRP-LDL interaction involves PCh-binding sites of CRP and the moieties on the LDL molecule such as apoB, cholesterol and PCh (Saxena et al., 1987;Nunomura & Hatakeyama, 1990;Bhakdi et al., 1999;Chang et al., 2002;Taskinen et al., 2002;van Tits et al., 2005).What seems surprising is that the blocking of the PCh-binding sites of CRP with phosphoethanolamine (PEt) may change the LDL-binding affinity Distinct effects of CRP isoforms on cell function of CRP.The binding of PEt-complexed CRP to enzymatically-modified LDL (E-LDL) was enhanced over that of uncomplexed CRP (Singh et al., 2008a).Furthermore, PEt-complexed CRP bound native LDL in whole serum (Singh et al., 2008b).The mechanism of PEt action on CRP remains unknown but it is possible that PEt might induce the formation of CRP aggregates, which are able to bind native LDL in vivo and thus prevent the development of atherosclerosis (Singh et al., 2008b;2008c).The interaction of CRP with lipoproteins may contribute to the regulation of LDL metabolism and foam cell formation in the arterial wall.As suggested, under normal conditions mCRP may exert its protective role via facilitating the safe clearance of retained native LDL from extracellular space, and thus lower the risk of formation of atherogenic LDL derivatives (e.g., ox-LDL).Under pathological conditions, mCRP can contribute to retardation of foam cell formation by reducing macrophage response to ox-LDL (Ji et al., 2006a).Recent findings also showed that CRP in its monomeric, but not native pentameric conformation, has the ability to bind a variety of immunoglobulins and other proteins (Boguslawski et al., 2007).
With regard to the blood or vascular cells, native CRP displays anti-inflammatory activity, whereas modified CRP exerts proinflammatory effects on shear-induced platelet adherence to neutrophils and neutrophil aggregation (Khreiss et al., 2004b), neutrophil-endothelial cell adhesion (Zouki et al., 2001), neutrophil survival (Khreiss et al., 2002a), and endothelium phenotype (Khreiss et al., 2004a).The regulation of cell function by CRP is due to CRP binding to and activation of several Fc receptors on human leukocytes and endothelial cells.CRP binding to the high affinity IgG receptor FcγRI (CD64) was first suggested by Muller and Fehr (1986), when it was shown that the binding of the protein to monocytic cells could be partially inhibited by monomeric IgG, and this finding was further supported by showing that the transfection of COS-7 cells with FcγRI increased CRP binding (Marnell et al., 1995).In the opinion of some authors (Bharadwaj et al., 1999;Stein et al., 2000), the low affinity IgG receptor FcγRIIa (CD32) is the major receptor for CRP on leukocytes, although the interpretation of data has been questioned by others (Saeland et al., 2001).As endothelium is regarded, CRP has been shown to bind mainly to CD32, but also to CD64 on HAECs (Devaraj et al., 2005).Preincubation with anti-CD32 or anti-CD64 antibodies inhibited maximal binding of CRP to HAECs, by 64% and 30%, respectively, whereas antibodies directed against FcγRIIIb (anti-CD16) had no effect.In line with those findings are functional studies on the effects of CRP isoforms on neutrophil function, which have revealed distinct Fcγ receptors for native and modified CRP on leukocytes.Unlike the native isoform, the modified CRP can bind with a low affinity to FcγRIIIb (CD16), thus inhibiting neutrophil chemotaxis (Heuertz et al., 2005).Native CRP is also able to induce the same effect, but the mechanism of nCRP interaction with neutrophils is distinct from that of mCRP since, for instance, the binding of human or rabbit native CRP was not affected by the presence of anti-CD16 antibody (Heuertz et al., 2005).Earlier reports have already described the involvement of distinct Fcγ receptors in the modulation of neutrophil function by various CRP isoforms.Moreover, some authors suggest even the opposite roles for native and mCRP in adhesion of neutrophils to HCAECs (Zouki et al., 2001), or platelets (Khreiss et al., 2004b) and neutrophil survival (Khreiss et al., 2002a).It has been proposed that the binding of nCRP to FcγRI and/or FcγRIIa is associated with shedding of l-selectin without neutrophil activation and the subsequent attenuation of polymorphonuclear leukocytes (PMN) adhesion to activated endothelial cells.In turn, when pentameric CRP is dissociated into its monomeric subunits, their binding to neutrophils is mediated via FcγRIIIb and results in an increased adhesion of PMNs to activated endothelium (Zouki et al., 2001).Again, the incubation of neutrophils with mCRP, but not nCRP, gives rise to delayed neutrophil apoptosis (Khreiss et al., 2002a).Some observations point that modified CRP can inhibit the development of morphologic features characteristic for apoptosis, as well as DNA fragmentation in the cells, and reduce the percentage of annexin V-positive neutrophils.The prolongation of neutrophil survival by mCRP is thought to be in part mediated by FcγRIIIb through stimulation of the PI 3-kinase/Akt and MEK/ERK signaling pathways, finally leading to inhibition of caspase-3, one of the key effectors of apoptosis (Khreiss et al., 2002a).
In addition to classical neutrophil recruitment into the inflamed endothelium, platelets bound to activated endothelial cells can interact with leukocytes, thus providing an indirect mechanism of neutrophil-endothelial interaction (Mine et al., 2001).Neutrophil rolling on platelets is modeled as a multiple-step process which involves initial tethering (mostly mediated by platelet P-selectin binding to P-selectin glycoprotein ligand-1 (PSGL-1) on leukocytes) followed by firm adhesion achieved by CD11b/CD18 (Mac-1) binding to GPIbα (Simon et al., 2000) and/or junctional adhesion molecule 3 (JAM-3) (Santoso et al., 2002).In studies on the effects of CRP conformers on the interplay between platelets and neutrophils, it has been found that modified CRP, in contrast to nCRP, may accelerate shear-induced platelet-neutrophil adhesion and neutrophil aggregation (Khreiss et al., 2004b).A complete abrogation of platelet-neutrophil adhesion and neutrophil aggregation required both anti-P-selectin and anti-CD18 antibodies, thus indicating the possible role of P-selectin and Mac-1 in the enhancement of adhesive interactions by mCRP.The observed functional antagonism between CRP isoforms in regulation of platelet P-selectin expression is also consistent with their action on platelet reactivity in response to various agonists (Fiedel & Gewurz, 1976a, 1976b;Fiedel et al., 1977Fiedel et al., , 1982;;Boncler et al., 2006).
Although the studies that evaluated the nCRP/mCRP functional differences have provided rather consistent results, they are limited by the fact that merely the recombinant form of mCRP (r m -CRP), and not native modified CRP isolated from biological material (b m -CRP) was used in the majority of those experiments.The recombinant and biological mCRP differ slightly in primary structure: r m -CRP compared to b m -CRP has two alanines instead of cysteine residues at positions 36 and 97, and an additional N-terminal formylmethionine residue.Besides, to enhance its solubility, r m -CRP is subjected to acylation (Khreiss et al., 2002a).These differences notwithstanding, these two mCRP forms showed high similarity in terms of biochemical characteristics, i.e., SDS/PAGE size, solubility, and antigenicity.With respect to in vitro activities, the effects of recombinant and biological mCRP on blood cells have not been extensively examined.Preliminary data published by Khreiss and coworkers (2002a), revealed a similar effect of r m -CRP and b m -CRP in delaying neutrophil apoptosis.

AnimAl modelS
First reports comparing the activities of CRP isoforms in animals appeared at the end of the previous century, when CRP was considered a therapeutically effective agent.Then, modified CRP, in contrast to native human CRP, was demonstrated to have a stimulatory effect on in vitro and in vivo murine thrombopoiesis (Potempa et al., 1996).In addition, in a murine model of mammary adenocarcinoma, it was shown that mCRP also has anticancer and antimetastatic activity, since the treatment with mCRP led to the inhibition of tumor growth and metastasis in mice (Kresl et al., 1999).
The most broadly available animal model, the mice, is considered largely useless regarding the study of CRP functions, because CRP is not an acute phase component in mice.To overcome this problem, a model of transgenic mice expressing human CRP (CRPtg mice) has been utilized to study the biological activities of human CRP in vivo.Unlike human CRP, mouse CRP circulates only in trace amounts, and its blood level does not change appreciably during inflammation (Szalai & McCrory, 2002).Furthermore, male CRPtg mice constitutively produce human CRP with serum levels that are considered to indicate high risk in humans.However, foreign antigenicity of human CRP in tissues of mice, despite their high homology (murine CRP shares 70% homology with human CRP), as well as the inability of human CRP to activate the complement pathway in mice, claimed by some authors studying effects of human CRP in mice, are regarded as serious limitations of the system and a possible source of contradictory results (Suresh et al., 2006).Even the most widely used apolipoprotein E knockout (ApoEˉ/ˉ) mouse model of atherosclerosis has been disapproved due to a not fully functional complement system in apoE-deficient mice (Reifenberg et al., 2005;Torzewski, 2005).With the use of wild type and human CRP-transgenic mice, Danenberg et al. (2003) described the first in vivo experiments of vascular injury, which clearly demonstrated a significantly faster and higher rate of arterial thrombosis in CRP-transgenic mice, thus pointing to a cause-effect relationship between CRP and thrombotic events.In addition, the influence of human CRP on monocyte-platelet aggregates (MPA) formation in CRPtg mice was examined (Danenberg et al., 2007).Following LPS injection, MPA levels were increased in CRPtg mice compared to wild-type mice, although no overt thromboses were observed.Transgenic expression of human CRP in apoE-deficient mice has been reported to accelerate (Paul et al., 2004), slow (Kovacs et al., 2007), or have no effect on atherosclerosis (Hirschfield et al., 2005;Trion et al., 2005;Tennent et al., 2008).The apoE-deficient transgenic mice have also been used to compare the effects of human nCRP and mCRP on the development of early atherosclerosis by using directly injected proteins (Schwedler et al., 2005).Those results evidently indicate that the configuration of CRP may determine differential effects on atherosclerosis in mice, as it does in humans.Native CRP enhanced plaque formation in apoEˉ/ˉ mice with early lesions, while mCRP reduced plaque formation and increased serum levels of the anti-inflammatory cytokine IL-10.Using the identical treatment protocols, those authors have also revealed that nCRP, but not mCRP, may impair endothelial function in aortic rings, which can in part be due to an increased activity of iNOS and peroxynitrite formation (Schwedler et al., 2007).In line with those findings is the paper by Bisoendial et al. (2007) showing a deteriorated endothelium-dependent vasodilator capacity and enhanced procoagulant response upon CRP challenge in patients with familial hypercholesterolemia compared to normolipidemic subjects.The pathophysiological significance of C-reactive protein in atherosclerosis has also been demonstrated in hypercholesterolemic rabbits.In comparison with Distinct effects of CRP isoforms on cell function normal rabbits, both cholesterol-fed and Watanabe heritable hyperlipidemic (WHHL) rabbits had significantly elevated CRP levels, and CRP concentration strongly correlated with the extent of atherosclerotic lesions in these animals (Sun et al., 2005).Taken together, the discrepancy in the outcomes of the animal studies does not allow definitive confirmation of the contribution of CRP in human atherogenesis.The development of a specific anti-CRP therapy, involving CRP inhibitors, would help in resolving this question.

ConTAminATionS in CRP PRePARATionS
The most commonly used commercially available preparations of CRP are recombinant CRP materials that most often contain a lot of contaminants, including bacterial lipopolysaccharide (LPS) and the preservative, sodium azide (NaN 3 ).Several studies have systematically evaluated the impact of different preparations of CRP on a variety of cell lines to determine whether inflammatory reactions are due more to CRP or to the contaminant.Van den Berg and coworkers ( 2004) have investigated the influence of three various CRP preparations (two azidefree CRP preparations, in-house recombinant CRP and ascites-derived CRP, versus commercial CRP) on vascular smooth muscle cell relaxation.They found that only the commercial CRP was able to induce vasorelaxation, which was further elucidated to be caused by the sodium azide present in this preparation and not by CRP itself.This data has been confirmed and extended by Swafford and coworkers (2005), who demonstrated no effect of CRP on regulation of vascular tone and proved that NaN 3 , as a source of NO, may induce vasodilatation, which is likely mediated by the cGMP-dependent activation of smooth muscle K + channels.In another series of experiments, changes in endothelial cell function (i.e., proliferation, morphology, apoptosis, expression of eNOS and ICAM-1, and levels of MCP-1, IL-8, or von Willebrand factor) in the presence of various CRP preparations, LPS, or azide have been monitored (Taylor et al., 2005).In accordance with previous observations (van den Berg et al., 2004), none of the reported effects of CRP on endothelial cells could be ascribed to CRP, but rather to the contaminants.The possible contribution of other non-CRP components of the solution, particularly sodium azide, has been analyzed with respect to the influence of CRP on proinflammatory mechanisms evoked by IL-1 β in cultured human vascular smooth muscle cells.Again, sodium azide, similarly to commercial CRP and vehicle medium, attenuated the induction of iNOS and NO release in this cell type.When sodium azide was removed from the CRP solution or the vehicle medium by dialysis, the previously observed effects were almost lost, which argued for sodium azide as the main mediator of such effects (Lafuente et al., 2005).The action of different CRP preparations -commercial recombinant CRP and human CRP purified from malignant ascitic fluid -were also tested in vivo.Intravenous injection into mice of malignant ascites-derived CRP did not induce an acute phase response of either mouse SAP or SAA.In contrast, injection of CRP from a commercial bacterial recombinant preparation induced a dramatic increase in the circulating concentration of both these proteins, compatible with the well known exquisitely sensitive murine acute phase response to endotoxin (Pepys et al., 2005).The recombinant material also triggered an inflammatory marker, TNFα, and signaling responses in mouse macrophages in vitro, whereas non-recombinant CRP did not, again suggesting that a contaminant of the commercial preparation is a probable cause of the inflammatory stimulation rather than CRP itself.That view, however, is not shared by all researchers, who showed that the biological effects of CRP on endothelium are specific to the protein and not related to contamination with either endotoxin or azide and occur at concentrations observed in patients to predict future vascular events (Verma et al., 2006;Dasu et al., 2007).
In summary, it seems reasonable to use in experiments the most pure preparation of CRP available; otherwise, adequate controls (including control buffers, control proteins, antibodies, blockers) are obligatory to minimize artifactual observations.On the other hand, the homogeneity of CRP preparations is equally important.For example, storage in the absence of calcium, complexed with cell membranes, or in a plastic container may lead to the formation of mCRP.In addition, prolonged storage is known to cause relaxation in the pentameric structure of CRP, which makes it more susceptible to the conversion into mCRP (Ji et al., 2006a).

fuTuRe PeRSPeCTiveS And ConCluding RemARKS
The role of CRP in pathophysiology/pathology has rarely been discussed with the reference to its structural heterogeneity.In the majority of papers CRP is considered a "villain", without even mentioning which form of this protein, pentameric or monomeric, we talk about.Some researchers may consider this fact justified, as far as there is no reliable commercial assays to well distinguish between the various biological forms of CRP.The available diagnostic methods enable determination of the whole pooled population of CRP molecules in the blood, which is of course an excessive simplification that may occasionally lead to inconsistencies in clinical reports.Regardless of these limitations, we have to remain aware that the biological (and evolutionary) roles of the native (pentameric) and modified (monomeric) forms of CRP are to a large extent opposite.Ignoring this fact may easily lead to accumulation of artefactual findings and/or detecting spurious relationships between CRP concentrations and its tentative biological effects.
Despite some methodological problems reported in the studies on CRP isoforms, there is accumulating evidence on the importance of pentameric and monomeric CRP in both the innate immune system and atherogenesis.Further investigations related to the development of a specific anti-CRP therapy are needed to definitely verify a direct pathogenic role of CRP and to better understand the role of CRP conformers in physiology and atherosclerotic lesion formation.It has been speculated that inhibition of CRP may be achieved with agents that directly affect CRP synthesis, action (transcriptional inhibition of CRP synthesis, antisense therapy, blockade of CRP-mediated complement activation, blockade of CRP receptors) (Scirica & Morrow, 2006) or influence CRP structure (conversion between pentameric and monomeric CRP forms, destabilization of CRP monomers) (Verma et al., 2004b).The first therapeutic inhibition of CRP with 1,6-bis(phosphocholine)hexane has been shown to be a promising approach to cardioprotection in rats (Pepys et al., 2006).Further studies with the use of anti-CRP agents should resolve the questions of the association between CRP and atherosclerosis and cardiovascular risk.