Joseph L. Witztum MD, University of California at San Diego, Basic Science Building-Room 1080, 9500 Gilman Drive, La Jolla, CA 92093, USA. (fax: 858 534 2005; e-mail: firstname.lastname@example.org).
During the oxidation of LDL, a central pathophysiological component of atherogenesis, a wide variety of chemical and physical changes occur leading to the generation of oxidation-specific neoepitopes. These epitopes are not only immunogenic, leading to adaptive humoral responses, but are also a prominent target of multiple arcs of innate immunity. The pattern recognition receptors (PRRs) of innate immunity are germ line encoded, conserved by natural selection, and bind to pathogen-associated molecular patterns (PAMPs) common on multiple structures. However, it is not intuitive as to why they should recognize oxidation-specific neoepitopes. Yet it is clear that multiple macrophage scavenger receptors, which are classic PRRs, recognize oxidation-specific epitopes, such as those found on oxidized LDL (OxLDL). Other innate proteins, such as C-reactive protein, also bind to OxLDL. Natural antibodies (NAbs), the humoral arc of innate immunity, provide a nonredundant role in the first line of defence against pathogens, but are also believed to provide important homeostatic house-keeping functions against self-antigens. Our work demonstrates that oxidation-specific epitopes, as found on OxLDL, are a major target of NAbs. In this review, we will discuss the specific example of the prototypic NAb T15/E06, which is increased in atherosclerotic mice and mediates atheroprotection, and discuss the potential role of NAbs in atherogenesis, and in inflammation in general. We also review data that oxidation-specific epitopes are generated whenever cells undergo programmed cell death, forming a common set of PAMPs recognized by oxidation-specific PRRs on macrophages, NAbs and innate proteins. We present the hypothesis that oxidation-specific epitopes on apoptotic cells exerted evolutionary pressure for the conservation of these PRRs and also serve to maintain the expansion of a substantial proportion of NAbs directed to these stress-induced self-antigens.
It is now widely appreciated that both adaptive and innate immunity impact atherogenesis, promoting atherogenesis, and providing atheroprotective influences as well (reviewed in [1–4]). There is an almost limitless number of T-cell and B-cell receptors (∼1014), generated through somatic mutations and induced junctional diversity at VDJ-gene recombination sites, which provide both cellular (via T cells) as well as humoral immunity [via secreted antibodies (Ab) from B-cell-derived plasma cells]. Whilst adaptive responses are delayed in time because of selection and maturation, once formed, they represent highly specific, high-affinity responses to newly recognized pathogens.
It is now clear that many different adaptive immune mechanisms can modulate lesion formation. T cells are prominent components of atherosclerotic lesions, and in general, they are believed to mediate proatherogenic responses. For example, γ-interferon-secreting Th1 cells in particular are considered proatherogenic, at least in murine atherosclerosis . In contrast, B cells, though not commonly found in lesions themselves, are present in adjacent adventitia and in draining lymph nodes [5–7]. Immunoglobulins, the product of such B cells, are prominently found in established lesions, in part bound to specific disease-related antigens, of which oxidation-specific epitopes associated with oxidized LDL (OxLDL) are prominent . Emerging evidence now indicates that overall, B cells convey an atheroprotective function and that the antibodies they secrete may directly modulate lesion formation . For example, B-cell-deficient LDLR−/− mice developed increased atherosclerosis , and splenectomy of apoE−/− mice, which decreased IgM anti-OxLDL Abs, resulted in increased atherosclerosis, an effect that could be rescued by infusion of B cells from aged apoE−/− mice . Furthermore, infusion of B cells into immunocompetent apoE−/− mice also conferred atheroprotection . In fact, we do not know which population of B cells conferred this protection and whether this was due to humoral immunity or conceivably to a newly emerging regulatory role of B cells as well . However, as the production of Abs is the most important function of B cells, it is highly likely that Abs play a major role in mediating this atheroprotection. The spleen is known to be the primary source of IgM Abs in uninfected mice [13–15] and our findings reviewed below suggest that the observed atheroprotective effects of B cells may in part be mediated by natural IgM Abs, the humoral arc of innate immunity.
In contrast to adaptive responses, innate immunity utilizes natural selection of receptors . The components of innate immunity are germ line encoded, and present at birth and/or matured via positive selection during the neonatal period or shortly thereafter. They are available for almost immediate defence against a perceived pathogen. They provide not only a vital and nonredundant role in the initial defence against invading pathogens , but are also postulated to play an equally important role in maintaining homeostasis against a variety of self-antigens [18–20]. Because these receptors are of limited numbers, they are focused on highly conserved motifs present on pathogens and self-epitopes [16, 21]. They are termed pattern recognition receptors (PRRs) and the conserved motifs to which they bind are termed pathogen-associated molecular patterns (PAMPs). The cellular compartment of innate immunity consists of monocyte/macrophages, dendritic cells and natural killer cells. Through a variety of endocytic PRRs, they bind and engulf pathogens bearing PAMPs, and through PRR-mediated signalling cascades, secrete cytokines and chemokines that provide both targeted and generalized responses that help orchestrate a coordinated defence against pathogens. By processing antigens engulfed by such PRRs, macrophages, dendritic cells and other antigen-presenting cells present epitopes to T cells, and thus serve as a vital link between innate and adaptive immunity.
In the context of atherosclerosis, it is apparent why adaptive immune responses would be recruited. As a result of the complex pathophysiology of atherogenesis, many alterations in structure and chemistry occur in the components found in the artery wall, generating novel epitopes and neo-antigens. For example, we have documented that oxidation of LDL leads to a wide variety of novel structures that are immunogenic, and form prominent antigens in the artery wall [8, 22–24]. We have termed these oxidation-specific epitopes and demonstrated that during the course of atherogenesis there is a profound adaptive humoral response, e.g. IgG, to a variety of oxidation-specific epitopes found on OxLDL [25, 26]. In murine and rabbit models, titres to such epitopes rise and fall with lesion progression and regression, respectively [26–28], typical of adaptive responses to exogenous infectious agents.
As noted above, innate immune responses utilize germ line-encoded receptors and effector proteins. Our work and others have now demonstrated convincingly that both cellular and humoral arcs of innate immunity are intimately involved in atherogenesis. Why should germ line-encoded PRRs, maintained by natural selection, recognize neoepitopes generated as a result of the atherogenic process? Yet it is now clear that PRRs do recognize common structural motifs prominently formed during atherogenesis. Indeed, we will review evidence that oxidation-specific epitopes constitute a prominent set of PAMPs recognized by multiple arcs of innate immunity and provide a rationale as to why such oxidation epitopes have provided the necessary selective pressure leading to recognition by multiple PRRs of innate immunity.
Oxidation-specific epitopes are PAMPs recognized by PRRs of innate immunity
Scavenger receptors of macrophages were so named because they bound and internalized modified LDL but not native LDL . Since the identification of the acetyl LDL receptor as the first such receptor, it is now appreciated that this receptor binds to a wide variety of other ligands, consistent with it being a classic PRR . Indeed, it is now recognized that there are a number of such scavenger receptors present on macrophages, and other cells, that recognize various oxidation-specific epitopes present on OxLDL, such as SRA-1,2 (the acetyl-LDL receptor), CD36, SR-BI, LOX-1, PSOX and others . Furthermore, evidence is beginning to accumulate that other innate PRRs also bind oxidation-specific epitopes. For example, there is recent evidence that members of the innate TLR family  also recognize such epitopes [33, 34]. Yet another example of an innate protein recognizing such epitopes is C-reactive protein (CRP). This protein was originally noted for its ability to bind to the phosphocholine (PC) adduct covalently bound to the cell wall polysaccharide (C-PS) of Streptococcus pneumoniae and for its ability to mediate enhanced clearance of this and other pathogens. However, we have shown that CRP also binds to the PC moiety of oxidized phospholipids (OxPL) present on OxLDL, but not native PC-containing phospholipids . These data demonstrate that a single PAMP, which shares molecular identity to an epitope found on oxidized lipids, is recognized by multiple arcs of innate immunity. As discussed below, oxidation-specific epitopes are an important target of natural antibodies (NAbs), the humoral arc of innate immunity.
B-1 cells secrete natural antibodies
Natural antibodies are an essential layer of innate immunity [19, 36, 37]. They are present at birth or shortly thereafter and can be found in gnotobiotic mice reared in the complete absence of external antigenic stimulation. Remarkably, they can also be enhanced later in life by positive antigenic selection. Because they are conserved by natural selection, the presumption is that fundamentally NAbs provide advantageous properties maintaining homeostasis, such as their crucial role in immediate host defences against pathogens [17, 37].
In mice, B cells can be broadly divided into conventional B-2 cells that mediate adaptive antibody responses, and B-1 cells that mediate the innate layer of humoral immunity and secrete NAbs which are predominantly IgM and IgA [37–39]. B-2 cells are continuously selected by somatic mutations and VDJ junctional diversity to meet newly encountered pathogens. In contrast, B-1 cells differ from B-2 cells in many ways, such as surface phenotype, anatomical location, restricted use of VH genes that are minimally edited and reflective of germ line usage, and importantly, their capacity for self renewal (reviewed in [20, 37–39]).
As noted above, the repertoire of B-1 cells is selected by natural conservation and is already established shortly after birth even in germ free mice . In uninfected mice, it is believed that most, if not all, IgM in plasma represent natural IgMs that are B-1 cell derived, and these are secreted predominantly from the spleen. However, established B-1 cell clones can be expanded later by antigen exposure leading to increased IgM levels in plasma [41–44]. Although the exact mechanisms leading to their selection and expansion are unclear, it is now appreciated that antigen selection during foetal and neonatal period leads to positive selection . As this appears to occur equally well in mice raised in germ free environments [46–48], this selection is of necessity made by endogenous self-antigens. This is in contrast to developing, self-reactive B-2 cells, in which such antigen encounter early in life leads to negative selection via apoptosis (=clonal deletion) or anergy. Thus, the repertoire of B-1 cells is selected to bind to evolutionarily important epitopes.
Oxidation-specific epitopes are evolutionarily important epitopes
Our laboratory first demonstrated the immunogenicity of OxLDL and that it generates a robust humoral and cellular response , findings that led to consideration of the epitopes of OxLDL as relevant disease-specific antigens in the atherosclerotic lesion. Remarkably, the immune response appears to be relatively specific to the oxidatively modified lipids and the oxidized lipid-apoB adducts formed during oxidation in a hapten-specific manner . Multiple copies of such oxidation-specific adducts may appear on the large LDL particle replete with polyunsaturated fatty acids (PUFA) susceptible to oxidation. Thus, OxLDL, containing multiple copies of oxidation epitopes on its surface, is capable of stimulating not only classic adaptive, but also thymus independent type 2 responses that typically require multivalent ligands [51, 52]. Thus, it is not surprising that oxidation of LDL, induced by various complex mechanisms [53–55], renders it immunogenic and that antibodies to a variety of oxidation epitopes are generated. For example, Abs to epitopes of malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE) , which are decomposition products of oxidized PUFA, and which form covalent adducts with amines of lysines or amino groups of lipids such as phosphatidylethanolamine (PE) are prominent. In addition, phospholipids containing a PUFA in the sn2 position can undergo decomposition generating OxPL, as for example, POVPC (1-palmitoyl-2-(5′-oxovaleroyl)-sn-glycero-3-phosphocholine) . These are but a few of the extensive modifications in both lipid and protein moieties that lead to the generation of neo-antigens, which we termed oxidation-specific epitopes . We have shown that these epitopes are present in atherosclerotic lesions of all species tested, as well as in humans [22, 24, 58]. In turn, Ab titres to these epitopes correlate with lesion progression and regression, at least in animal models [26, 28]. Thus, immunologically recognized OxLDL is present wherever atherosclerotic lesions are present, and a humoral immune response occurs as well. Most importantly, the presence of oxidation-specific epitopes is not only restricted to OxLDL and atherosclerotic lesions, but in fact these epitopes are ubiquitously present in inflamed tissues found in many different diseases, including rheumatoid arthritis, Alzheimer’s disease  and haemachromatosis  amongst others. In particular, as described below, because oxidative events are a key process in cells undergoing apoptosis, oxidation-specific epitopes are universally present on the cell surface as well as in apoptotic bodies – often termed microparticles – that bud off from such cells [61–63]. Such microparticles are found in plasma and have many bioactive and proinflammatory properties. Because the generation of apoptotic cells and inflammatory reactions take place throughout life, we suggest that oxidation-specific epitopes associated with these processes are important self-epitopes and constitute a dominant class of PAMPs that confer a strong evolutionary pressure driving PRR selection of innate immunity.
Many anti-OxLDL are IgM natural antibodies
Cholesterol-fed apoE−/− mice have high autoantibody titres, in particular IgM, to a wide variety of oxidation-specific epitopes, which enabled us to clone a panel of IgM B-cell hybridomas from the spleens of these mice with specificity for OxLDL or MDA-LDL . Amongst these, we cloned seven IgM monoclonals that all had specificity for OxLDL and although we thought each was unique, we subsequently learned that each was identical in its variable region antigen-binding site. Thus, the prototypic E06 bound to both the lipid and apoB moiety of OxLDL, and specifically, to the PC headgroup of OxPL, (such as POVPC) but not to the PC of native PL . Importantly, E06 inhibited the uptake of OxLDL by macrophage scavenger receptors CD36 and SR-BI [65, 66], as did POVPC linked to a peptide. Indeed, the PC moiety of OxPL was sufficient itself to mediate binding of OxLDL to CD36 . These data demonstrate that the PC of OxPL in OxLDL mediated not only the binding of E06, but also was a ligand mediating binding of OxLDL to scavenger receptor CD36. Thus the PC of OxPC is a PAMP recognized by two PRRs of innate immunity. Because all the cloned antibodies binding to OxLDL were IgM, which are thought in large part to represent NAbs in uninfected mice, we sequenced VH / VL chains from four hybridomas secreting IgM to OxLDL , which revealed them all to be 100% homologous through 350 bp in both their VH / VL genes, as well as 100% homologous to the classic germ line-encoded NAb T15, secreted by a well-characterized B-1 cell clone described over 30 years ago . In fact, E06/T15 is considered to be the classic germ line NAb. The original T15 clone was a spontaneous murine plasmacytoma that secreted an IgA with the T15 idiotype. It binds to PC covalently linked to the C-PS of pathogens, i.e. the PC is not part of a phospholipid. T15 confers optimal protection to mice from lethal infection with S.pneumoniae . Thus, there is molecular (and immunological) mimicry between the PC of OxPL present on OxLDL and the PC present on the cell wall of pneumococcus and many other infectious pathogens. This dual specificity has been described as a characteristic of NAbs .
The fact that the same T15 NAb that blocked the uptake of OxLDL by macrophages also bound the PC of common microbial pathogens suggested that T15 natural IgMs might ameliorate atherosclerosis. This hypothesis was confirmed by the demonstration that immunization of cholesterol-fed LDLR−/− mice with heat-inactivated PC-containing pneumococci, which induced high titres of anti-OxLDL IgM (predominantly of the T15 clonotype), reduced atherosclerosis as measured at the aortic valve . Plasma of these mice had an enhanced ability to inhibit the uptake of OxLDL by macrophages. The anti-atherogenic role of T15/E06 IgM Abs was corroborated by Faria-Neto et al., who demonstrated that direct injection of T15-IgM inhibited vein graft atherosclerosis . Furthermore, Caligiuri et al. immunized apoE−/− mice with PC-KLH, using the heterologous protein KLH to induce an immune response and also observed a reduction in atherosclerosis . In their experiments, IgG responses were also induced along with IgM. They also observed that the immune antiserum was capable of inhibiting the uptake of OxLDL by macrophages, suggesting that IgMs of the T15/E06 idiotype were expanded.
In analogy with the previously described approach, we generated another set of monoclonal Abs from the spleens of nonimmunized cholesterol-fed LDLR−/− mice . Remarkably, we once again identified only IgM Abs that had specificity for different oxidation-specific epitopes. One particular IgM NAb cloned – LRO1 – was shown to have specificity for oxidized cardiolipin and shown to bind to its cognate epitope in atherosclerotic lesions. This demonstrates the existence of yet another natural IgM Ab that is expanded in atherosclerotic mice and has specificity for oxidation-specific epitopes. In ongoing studies in our laboratory, we have cloned a variety of other NAbs that have specificity for other oxidation-specific epitopes, for example, to structures formed when MDA is adducted to the epsilon amine of lysines. Such NAbs appear to be very prominent in the NAb repertoire of mice.
B-1 cells can be regulated in vivo
Whilst NAbs with certain specificities, such as those directed to PC, can be induced by antigenic stimulation (see above), NAb in general are believed to remain at tightly regulated levels. Nevertheless, a variety of factors are postulated to modulate B-1 cell function, and thus IgM secretion, but in a noncognate manner. For example, B-1 cells constitutively express the IL-5 receptor and IL-5 transgenic mice have elevated levels of serum IgM and IgA . We showed that IL-5 stimulates anti-OxLDL IgM and T15/E06 IgM specifically in vitro and in vivo . Further, we generated LDLR−/− mice deficient in IL-5 by transfer of bone marrow from IL-5−/− mice. The cholesterol-fed IL-5−/−LDLR−/− chimeric mice had lower titres of T15 clonotypic Abs and significantly more atherosclerosis than IL-5 competent controls . These data indicate that IL-5 is an important factor for the natural expansion of T15/E06 IgM during atherogenesis. There are also other factors that have been described to stimulate NAb production by B-1 cells, e.g. certain TLR ligands, though the role of these factors in vivo and in the context of atherosclerosis remains to be described .
Functional roles of oxidation-specific IgM natural antibodies
A number of potential protective mechanisms for NAbs can be proposed . As discussed above, in vitro studies have shown that T15/E06 Abs can inhibit the scavenger receptor-mediated binding and uptake of OxLDL by macrophages and in vivo expansion, whether by immunization or direct infusion, also is anti-atherogenic [69–71]. In addition, we have shown that T15/E06 IgM Abs are also able to inhibit the activation of endothelial cells induced by apoptotic cells and apoptotic blebs carrying oxidation epitopes [62, 76]. Thus, T15/E06 IgM Abs have the capacity to neutralize the biological activity of OxPL of apoptotic cells, OxLDL and minimally modified LDL (which all display PC-containing OxPL). By analogy, different oxidation-specific natural IgM Abs may act by neutralizing the many documented pro-atherogenic and pro-inflammatory effects of other components of OxLDL.
Finally, we also demonstrated the formation of circulating IgM–apoB immune complexes and T15–apoB immune complexes in plasma of hypercholesterolaemic mice . Although the presence of high titres of anti-OxLDL IgM in apoE−/− mice did not seem to accelerate the clearance of injected OxLDL , one could still speculate that the LDL particles bearing oxidation epitopes would be redirected to clearance by reticuloendothelial organs, such as spleen and liver, which would decrease the entry of potentially atherogenic LDL particles into the artery. Indeed, we showed that Abs to nonenzymatically glycated LDL redirected glycated LDL in the circulation away from the artery and into spleen, bone marrow and liver . Taken together, many potential atheroprotective functions of natural IgM Abs have been demonstrated in vitro, but a detailed identification of their functional capacities in atherosclerosis awaits further study.
Oxidation-specific NAbs bind to apoptotic cells
Natural antibodies are selected by evolution. The observation that E06 binds to infectious pathogens suggests that such pathogens could be one class of selecting agents leading to the conservation of reactive B-1 cell clones. However, as alluded to above, our work has uncovered another set of self-antigens, which we postulate might be the fundamentally most important selecting antigens for oxidation-specific NAbs as well as other innate PRRs. Cells undergoing programmed cell death develop mitochondrial disruption, leading to generalized enhancement of oxidative events. In turn, a wide variety of lipids are oxidized, including both phosphatidylserine and phosphocholine. We  and others [79, 80] have shown that there is a greatly enhanced content of OxPL in apoptotic cells. In addition, such epitopes are greatly enriched in the apoptotic blebs that bud off from such cells [61, 62] and are present in the circulation as microparticles. Apoptotic cells are immunogenic and proinflammatory if not properly cleared  and thus, there would be evolutionary pressure to provide mechanisms to efficiently clear such damaged and potentially pathogenic agents. Indeed, we have demonstrated that all our oxidation-specific monoclonal antibodies also bind to apoptotic cells and apoptotic bodies [61, 72, 76]. Further, the oxidation-specific epitopes are functionally important ligands mediating binding and uptake of apoptotic cells by macrophages, mediated by scavenger receptors such as CD36, i.e. they are PAMPs recognized by macrophage PRRs. Indeed, E06 can block the uptake of apoptotic cells by macrophages, as can the synthetic E06 epitope POVPC linked to BSA [35, 76]. In support of this, we have shown that the POVPC peptide can directly bind to CD36 [65, 67].
Because such oxidation-specific IgMs bind to apoptotic cells, they might well facilitate the enhanced clearance of such dying cells. Indeed, Ogden et al. have demonstrated that in the presence of complement, E06/T15 has the capacity to enhance apoptotic cell clearance in vivo . One could speculate that NAbs to a variety of different oxidation-specific epitopes would enhance clearance of cells expressing such epitopes resulting from apoptosis or oxidative damage. Similarly, they would facilitate the clearance of the myriad microparticles that are formed from damaged and dying cells. Failure to adequately clear the daily burden of apoptotic cells would likely lead to inflammation and even autoimmunity. Indeed this appears to be the case in many murine models of lupus.
Figure 1 summarizes our general hypothesis that oxidation-specific epitopes are an important target of innate immunity, and form a novel set of PAMPs that are recognized by multiple PRRs of innate immunity. Thus, the PC of OxLDL and of apoptotic cells is recognized by specific NAbs, such as E06, by scavenger receptors, such as CD36 and SR-B1, and by the innate protein CRP. In turn, we suggest that each of these PRRs will also recognize the PC present on the C-PS of pathogens, such as S. pneumoniae. Similarly, we predict that other oxidation-specific epitopes, such as MDA, 4-HNE and oxidized cardiolipin, will be PAMPs for other NAbs, scavenger receptors and innate proteins. We predict that other types of posttranscriptional modifications, such as advanced glycation end products will similarly be found to be targets of these innate PRRs.
There is now ample evidence that innate immunity is critically involved in the pathogenesis of atherosclerosis, as documented by an array of studies in mouse models of atherosclerosis. NAbs are an essential layer of innate immunity  and at least in the selected situations reviewed above, appear to play an anti-atherogenic role in murine atherosclerosis. Based on their properties to bind both microbial antigens as well as oxidation-specific self-antigens, as present on OxLDL and apoptotic cells, they undoubtedly not only play an important role in immediate host defence against microbial infections, but importantly, homeostatic house-keeping functions by preventing the deleterious accumulation of stressed self . Indeed, most NAbs have been hypothesized to contain such dual specificities  and our evidence suggests that oxidation-specific epitopes constitute a major class of antigens to which they bind. Because NAbs have been conserved by natural selection, they must be of benefit to the host in a general context. We postulate that defining oxidation-specific epitopes to which such NAbs bind may lead to the identification of antigens capable of stimulating anti-inflammatory and atheroprotective innate immunity, at least in mice. Future studies are needed to test this hypothesis and to identify the functional properties of such NAbs, both in health and in disease states such as atherosclerosis and inflammation. In addition, we need to define whether such NAbs exist in humans and to determine their functional significance. It can be envisioned that this understanding might ultimately identify novel therapeutic strategies to interfere with atherogenesis and inflammatory states in general.