Correspondence Runlin Han, Research Center of Plasma Lipoprotein Immunology, College of Animal Medicine, Inner Mongolia Agricultural University, 306 Zhaowuda Road, Huhhot, 010018, China. Tel./fax: +86 471 4310523; email: firstname.lastname@example.org
Plasma lipoproteins (VLDL, LDL, Lp[a] and HDL) function primarily in lipid transport among tissues and organs. However, cumulative evidence suggests that lipoproteins may also prevent bacterial, viral and parasitic infections and are therefore a component of innate immunity. Lipoproteins can also detoxify lipopolysaccharide and lipoteichoic acid. Infections can induce oxidation of LDL, and oxLDL in turn plays important anti-infective roles and protects against endotoxin-induced tissue damage. There is also evidence that apo(a) is protective against pathogens. Taken together, the evidence suggests that it might be valuable to introduce the concept that plasma lipoproteins belong in the realm of host immune response.
plasminogen binding group A streptococcal M-like protein
trypanosome lytic factor
tissue plasminogen activator
urokinase-type plasminogen activator
very low-density lipoprotein
A large body of evidence suggests that plasma lipoproteins may play an important role in host defense as a component of the innate immune system (1–3). Some of the viewpoints in these reviews are briefly summarized below. Infection and inflammation induce an APR with multiple alterations in lipid and lipoprotein metabolism. Briefly, in nonhuman primates and humans, VLDL concentrations are increased whereas LDL is either unchanged or decreased and HDL is decreased (For a review, see reference 3). Lipoproteins, including VLDL, LDL, lipoprotein(a) and HDL, can detoxify LPS from Gram-negative bacteria and LTA from Gram-positive bacteria, HDL being the most potent of these lipoproteins (2). The LPS- neutralizing role of HDL has been extensively studied both in vitro and in vivo (for reviews, see references 4, 5). Moreover, infusion of rHDL can blunt endotoxin-induced procoagulant activation in humans (6). Plasma HDL concentrations are significantly lower in patients with cirrhosis than in controls. Recently, an ex vivo study in which rHDL was incubated with either whole blood or monocytes from patients with severe cirrhosis showed that rHDL can significantly decrease LPS-induced overproduction of proinflammatory cytokines (tumor necrosis factor alpha and interleukin-10). This suggests that administration of rHDL might be protective against LPS-induced inflammation in cirrhosis patients (7).
S. aureusα-toxin is bound, and almost totally inactivated, in vitro by LDL (8). Hypocholesterolemic rats have been shown to have an increased endotoxin-induced mortality which can be ameliorated by injecting purified human LDL (9). In accordance, much epidemiological and clinical evidence suggests that high cholesterol may protect against respiratory and gastrointestinal infections (1).
Lipoproteins bind and neutralize a vast number of enveloped and non-enveloped DNA and RNA viruses. VLDL and LDL particularly protect against certain types of virus, such as togaviruses and rhabdoviruses, whereas HDL shows broader antiviral activity (3). Trypanosomes are unicellular parasites that cause sleeping sickness in animals. However, Trypanosoma brucei do not cause infection in humans because TLF, a subfraction of native human HDL containing haptoglobin-related protein, apolipoprotein L-I and apoA-I, protects against trypanosome infection (10).
Infection and inflammation cause oxidation of LDL (11). One of the oxidized components of oxLDL, oxPAPC, inhibits LPS-induced adhesion of neutrophils to endothelial cells and serves a protective role in endotoxin-induced tissue damage (12). In addition, both minimally oxidized LDL (13) and oxPAPC (14, 15) have been demonstrated to inhibit LPS signaling, consequently detoxifying LPS.
However, the anti-infective roles played by plasma lipoproteins have not been studied systematically. Here, we add some novel and/or relevant evidence and provide a theoretical analysis to support the role of plasma lipoproteins in the host defense system.
THE COMPLEX INTERPLAY BETWEEN HDL AND INFECTIONS
Leishmania can cause human leishmaniasis ranging from mild cutaneous and mucocutaneous lesions to fatal visceral infections. Recently, TLF was demonstrated to inhibit Leishmania-induced intracellular infection in TLF-transgenic mice by directly damaging Leishmania parasites which have proliferated in the phagolysosomes of macrophages, within which TLF accumulates and is activated under acidic conditions. However, TLF mice are not protected against infection by T. cruzi, which invades many cell types and transiently passes through phagolysosomes (16). The anti-infectious role of HDL may be attenuated by serum opacity factor, a virulence factor of class II M types of GAS, because it opacifies mammalian sera by binding and disrupting HDL (17).
HDL also has antiviral activity (18–20). A large number and variety of virus species can be inhibited from entering host cells by HDL. In the last decade studies have focused on the interplay between HDL and the influenza and hepatitis C viruses. D-4F, an apoA-I mimetic peptide, has been reported to have antiviral activity in vivo, whereas infection may cause HDL to lose its anti-inflammatory properties (21). In contrast, HDL may enhance HCV infection through the scavenger receptor class B type I (22–24). Another adverse effect is that infusion of reconstituted HDL can enhance the growth of Candida albicans (25). In addition, a recent report that apoA-I can neutralize LPS, whereas apoA-II may suppress this activity, further complicates the issue (26).
THE COMPLEX INTERPLAY BETWEEN LDL AND INFECTIONS
Netea and coworkers conducted a series of experiments to test whether LDLR−/− and apoE−/− mice are protected against lethal endotoxemia and severe Gram-negative infections (27–29). The results showed that the LDLR−/− mice have anti-Klebsiella pneumoniae capabilities due to an increase in endogenous LDL concentration (27), whereas apoE−/− mice, which have a great increase in VLDL and a lesser increase in LDL concentrations, are more susceptible to the infection than normal mice (28). These results indicate that the absence of apoE itself in the knock-out mice might abolish the K. pneumoniae LPS-neutralizing effect of lipoproteins, although high cholesterol concentrations are generated in these mice (28). However, a recent report shows that knockout of either apoE or the LDLR gene in mice results in a much higher resistance to infection by Salmonella typhimurium than is found in control mice (29). This may suggest a novel LDL-related anti-infective mechanism, other than detoxification of LPS, in which direct LDL-bacterial cell interaction might be involved (29).
Other instances of direct bacterial cell-LDL binding have also been reported. LDL can bind to pH6-Ag, a potential adhesive for the bacterial pathogen Yersinia pestis (30). We have also reported that, in M1, M12, M28 and M41 strains of GAS, apoB100-containing lipoproteins, mainly LDL, can bind to Scl1 (group A streptococcal collagen-like protein 1), which is a cell surface virulence factor of GAS (31). However, the pathophysiological significance of these LDL-pathogen interactions remains to be elucidated by more experiments including animal experiments. In addition, it may be valuable to test whether LDL is able to bind cells of other pathogenic bacteria.
Moreover, apoE−/− mice are more susceptible than control mice to Listeria monocytogenes (32) and Mycobacterium tuberculosis (33). Intraperitoneal pretreatment with LDL blocks Vibrio vulnificus LPS-induced lethality in mice (34). Therefore, the relationship between the anti-infective spectrum of lipoproteins (LDL and HDL) and LPS structure in different species of Gram-negative bacteria is highly complex. A recent study confirmed that LDL or HDL detoxify E. coli LPS more strongly than do whole E. coli or Neisseria meningitides (35). In contrast, in both apoE−/− and LDLR−/− mice, hypercholesterolemia exacerbates virus-induced immunopathological liver disease (36) and decreases resistance to C. albicans (37, 38). These findings imply that neither apoE-containing lipoproteins nor LDL provide protective activity against the above two pathogens. However, the mechanism is unclear.
ANTI-INFECTIVE ACTIVITY OF OxLDL
OxLDL is able to up-regulate the expression of some scavenger receptors (Classes A and E) in macrophages; consequently it promotes phagocytosis of Gram-positive and Gram-negative bacteria and endotoxin clearance (39). In addition, the DNA of many pathogens which has been detected in atherosclerotic plaques (40, 41) may be due to bacterial ingestion by macrophages, on which some oxLDL-inducible scavenger receptors are highly expressed. Recently, it has been shown that infection causes an increase in oxidation of LDL and that oxLDL, in turn, promotes phagocytosis of the responsible pathogens (42).
Notably, the natural antibody against PC elicited by oxLDL (43, 44) might provide a broad protection against pathogens since PC has been detected on a number of prokaryotes including Gram-negative and Gram-positive bacteria (such as S. pneumoniae, Bacillus spp., Clostridium spp., H. influenzae, N. meningitides, N. gonorrhoea, Pseudomonas aeruginosa, alpha-hemolytic streptococci, Actinobacillus actinomycetemcomitans, and T. cruzi), and parasites (such as the protozoan Leishmania major, trematodes, tapeworm, Diphyllobothrium latum, some gastrointestinal nematodes and all filarial nematodes), as well as many fungi (45–49). It has been repeatedly demonstrated that anti-PC antibody helps to prevent upper airway infections caused by S. pneumoniae and H. influenzae (50–53). The anti-PC antibody can also protect against A. actinomycetemcomitans in vitro (54). In contrast, disadvantages for the host in expression of PC by pathogens have also been reported, since PC mediates persistence and invasiveness (48). For example, surface PC appears to contribute to the persistence of H. influenzae in the respiratory tract (55, 56). The above contradictory findings might be explained by the difference in molecules to which PC is attached. On the one hand, PC is only a hapten, while on the other hand anti-PC antibodies usually recognize the PC epitope in oxidized phospholipids (57, 58). There have been no systematic studies on the anti-infective activities of anti-PC antibody so far. Thus, we suggest that it is necessary to extend the study of anti-infective properties of the anti-PC antibody to other pathogens. In addition, in vitro study has shown that oxLDL, but not native LDL, is a potent inhibitor of cell entry for a broad range of HCV strains (59) and for Plasmodium sporozoite (60).
Therefore, oxidation of LDL may be an important response of the host against infections. OxLDL induced by infection may initially protect the host from the harmful effects of bacteria, viruses, and parasites. However, the prolonged presence of oxLDL may contribute to atherosclerosis, since it is thought to play a central role in atherogenesis (61). Thus it obeys the rule that immunity is a double-edged sword (12). Furthermore, in addition to PC, other neoepitopes of oxLDL also induce production of anti-oxLDL antibodies, which may play an important role in the modulation of atherogenesis (62).
In contrast, an in vitro study has shown that mildly oxidized LDL inhibits production by human peripheral blood mononuclear cells of antibody against C. albicans (63). However, this has not been demonstrated in vivo. Thus, it would be valuable to further investigate the anti-infective spectrum of oxLDL.
POSSIBLE ANTI-INVASIVE ACTIVITY OF Lp(a)
Lp(a) has been studied extensively since it was first described 40 years ago (64–67). Lp(a) is present in humans, Old World nonhuman primates, and the European hedgehog. Lp(a) is a cholesterol-rich lipoprotein similar in structure to LDL, but containing an additional apo(a) moiety bound to apoB100 via a disulfide bridge. Apo(a) shares a high homology with Plg. It has been reported that Lp(a) is a risk factor for cardiovascular disease via both atherogenic and thrombotic mechanisms because Lp(a) contains LDL and there is a similarity between apo(a) and Plg. However, the physiological role of Lp(a) remains elusive. Some authors suggest that Lp(a) might promote or accelerate wound healing and repair of tissue injuries (68). Apo(a), independent of its interaction with Plg, has recently been demonstrated to inhibit neutrophil recruitment in some peritoneal inflammatory models of apo(a) transgenic mice. This suggests that apo(a) might play a beneficial role in humans by suppressing inflammation (69).
It has been reported that a vast number of pathogens, including a number of Gram-negative and Gram-positive bacteria, recruit human plasmin(ogen) through the lysine residue in their surface Plg receptors, and some even produce Plg activators to penetrate tissue barriers (70). Activation and regulation of plasmin(ogen) may play an important role in the invasion of pathogens (71, 72). For example, streptokinase secreted from GAS is highly specific for activation of human Plg, and human Plg-transgenic mice are highly susceptible to GAS (73). Additionally, the PAM-positive M type 53 of GAS causes a 60% increase in mortality in transgenic mice when compared with control mice, whereas the PAM-negative isogenic mutant of this strain shows only minimal virulence in both wild-type and transgenic mice. Therefore, many infections may be inhibited or prevented to some extent if recruitment and activation of host Plg by pathogens is blocked or inhibited (73). We consider that Lp(a) might constitute part of the host defense system for inhibiting pathogens by recruiting host plasmin(ogen), since it has an anti-fibrinolytic activity (74). In fact, one in vitro experiment has shown that Lp(a) inhibits Plg activation by streptokinase (75). We have also reported that Lp(a) inhibits binding between S. aureus CMCC26003 and Plg, and subsequently decreases urokinase-type Plg activator-activated fibrinolytic activity on streptococcal cells in vitro (presented at the Ninth National Conference on Plasma Lipoprotein, Xining, China, October 24, 2008). Thus, further research on the interaction of Lp(a) with other pathogens is warranted.
In addition, Lp(a) is as potent as LDL in inhibiting LPS-stimulated tumor necrosis factor synthesis by human mononuclear cells (76). This implies that Lp(a) may be an important factor in determining the amplitude of the response to LPS in humans, as there is a great variation in Lp(a) concentrations among individuals (76). Therefore Lp(a), with its combination of apo(a) and LDL, might be a potent anti-infective molecule in humans.
Taken together we suggest that, besides their primary role in lipid transport, lipoproteins participate in innate immunity since they have broad preventive effects against bacterial, viral and parasitic infections (summarized in Fig. 1 and Table 1). Many experiments have been conducted using mouse models. However, the effects of infection and inflammation on cholesterol metabolism differ between primates and rodents (3); in the latter HDL seems to have the same role as LDL in humans and vice versa (77). Therefore, data from studies with mouse models may not be fully applicable to the situation in humans, but plasma lipoproteins are doubtlessly important components of the immune system in both groups. Lipoproteins are complicated supermolecular complexes and their multi-functional characteristics warrant further experiments and clinical trials.
Table 1. The complex interplay between plasma lipoproteins and infections
a, including oxLDL and oxLDL-elicited anti-PC antibody; h, human lipoprotein(s) are involved; m, animal study with mice; n, an anti-infectious of the lipoprotein has not been observed or reported; t, in vitro study; v, in vivo study; –, the lipoprotein enhances infection by the pathogen; +, the lipoprotein protects against infection by the pathogen; ?, further studies to test whether the lipoprotein is a friend to, or foe of, the pathogen are required.
In addition, the cumulative evidence suggests that infections may induce or promote atherosclerosis (21, 41, 78–82). From our point of view, scientists might readily explain the correlation between infections and atherosclerosis on the basis of the double-edged sword property of the anti-infective immune effects of lipoproteins, and might be able to design more rational and systemic experiments to test this correlation.
CONFLICTS OF INTEREST
The author declares that there are no competing financial interests.
This work was supported by a start-up grant from the Inner Mongolia Agricultural University [Grant number K32615]. A part of the abstract is published in Journal of Immunology, 2009, 182: 134.74 as a meeting abstract (http://jimmunol.org/cgi/content/meeting_abstract/182/1_MeetingAbstracts/134.74). We apologize to the many investigators whose work could not be cited directly because of space considerations.