LBP [LPS (lipopolysaccharide)-binding protein] was discovered approximately 25 years ago. Since then, substantial progress has been made towards our understanding of its function in health and disease. Furthermore, the discovery of a large protein family sharing functional and structural attributes has helped in our knowledge. Still, key questions are unresolved, and here an overview on the old and new findings on LBP is given. LBP is an acute-phase protein of the liver, but is also synthesized in other cells of the organism. While LBP is named after the ability to bind to LPS of Gram-negative bacteria, it also can recognize other bacterial compounds, such as lipopeptides. It has been shown that LBP is needed to combat infections; however, the main mechanism of action is still not clear. New findings on natural genetic variations of LBP leading to functional consequences may help in further elucidating the mechanism of LBP and its role in innate immunity and disease.
Discovery of LBP
Historically, LBP [LPS (lipopolysaccharide)-binding protein), the focus of this review, and BPI (bactericidal/permeability-increasing protein) were the first members of the family of proteins with a BPI/LBP/PLUNC (palate, lung and nasal epithelium clone)-like domain discovered. Both proteins were first described in the 1980s, and LBP was first isolated from rabbit serum by Peter Tobias and Richard Ulevitch . It was named LBP for its ability to bind to LPS, a characteristic cell wall compound of Gram-negative bacteria. BPI, mainly found in the granules of neutrophilic granulocytes, shares the ability to bind to LPS and, as early as 1988, the existence of a protein family was proposed . LBP binds to the highly conserved lipid A portion of LPS and strongly modulates its immunostimulatory capacity [3,4]. Cloning of rabbit and human LBP revealed a high degree of sequence homology with human BPI; however, the functions of the two differ strongly, as LBP can enhance LPS-mediated effects on immune cells, whereas BPI acts inhibitorily [4,5]. Parallel with the cloning and first functional analysis of LBP, it was found that CD14 acts as an acceptor of LPS both as phosphatidylinositol-linked form on cellular surfaces and as soluble protein [6,7]. It took approximately 10 more years to discover how cellular recognition of LPS mediated by LBP and CD14 could lead to signal transduction. We know now that, for this, a presentation of LPS to homodimers consisting both of one MD-2 and one TLR4 (Toll-like receptor 4) molecule is required [8–13]. This quite complex LPS recognition apparatus may reflect the importance of an early and controlled response to LPS from Gram-negative bacteria. Within this cascade of LPS recognition molecules LBP is the only one with the ability to recognize LPS multimers formed immediately after LPS release into the bloodstream. LBP apparently monomerizes LPS, and only this monomer can be seen by other LPS acceptors, which are CD14, MD-2 and TLR4 [14,15].
The family of proteins with a BPI/LBP/PLUNC -like domain
Both, human LBP and BPI are located in close proximity on chromosome 20 and share an intron–exon pattern with other human lipid-binding proteins, which was later also found to be characteristic of the entire BPI/LBP/PLUNC-like domain family [16–19]. Two human serum proteins, PLTP (phospholipid-transfer protein) and CETP (cholesteryl ester-transfer protein), are important regulators of lipoprotein levels in humans and have received great attention as potential therapeutic targets in cardiovascular medicine [20,21]. Similarly to LBP (see below), within these proteins frequent mutations have also been described and linked to disease susceptibility [22,23]. Starting approximately 10 years ago, with the family of proteins expressed in PLUNC, a large number of other structurally and potentially functionally related proteins have been described [24,25]. Their genomic organization was found to be similar to the BPI–LBP–CETP–PLTP family, and their genes are also located on chromosome 20 [26,27]. First studies focusing on polymorphisms in PLUNC genes have been performed and a relevance of disease susceptibility also has been proposed lately . Only for BPI has the three-dimensional structure been solved; however, computer-generated models for the other family members revealed the existence of similar functional domains [29–31]. For LBP, the LPS-binding domain has been clearly identified, and the homologous regions probably also have similar functions in related proteins, as has been shown by the generation of fusion proteins and mutants lacking this domain [32–34]. Members of the BPI–LBP–CETP–PLTP–PLUNC protein family have been identified in many different animal species, and a phylogenetic analysis has revealed a rapid evolution of this protein family .
While LBP has been identified early in the mouse leading to the generation of knockout mice allowing for detailed functional analyses (see below), for BPI for many years the murine gene could not be identified . However, studies have clearly identified murine BPI [36a]. LBP has been found in many species; however, a recent analysis of the chicken genome failed to identify an LBP homologue, whereas homologues for many of the other PLUNC members were found . Chickens are known to be LPS-resistant and their flora is significantly different from the human microbiota. Further research may elucidate whether LBP has a role in controlling the diversity of the commensal flora.
Functions of LBP as an acute-phase protein
LBP is mainly synthesized in the liver and is induced and released as type-1 acute-phase protein [38–40]. IL (interleukin)-1, IL-6 and dexamethasone synergize in LBP induction and transforming growth factor-β inhibits its release . Other sources of LBP are epithelial cells of the lung, the GI (gastrointestinal) space, the kidney and the reproductive tract [42–46]. Particularly within the lungs, a rise in LBP on treatment with LPS may be mainly caused by local LBP synthesis as has been shown recently [47,48]. Assessment of LBP serum levels has been suggested as a useful marker during systemic infectious complications such as sepsis and septic shock [49,50]. Compared with other acute-phase proteins, however, the rise of LBP is relatively slow, with a maximum 2–3 days after the onset of the acute phase. High levels of LBP in serum surprisingly have been shown to inhibit LPS-induced inflammation [51–53]. This effect is dependent on lipoproteins or chylomicrons, which may act as ‘acceptor’ of LPS [54,55]. Mice could be saved during experimental septic shock by the intraperitoneal injection of recombinant LBP . Here it is shown that mice are protected from liver injury, and recently a specific role for LBP in liver-related inflammation has been suggested . Very high levels of LBP are seen in acute inflammatory diseases involving the GI region such as pancreatitis . It has also been shown that, within the lungs, LBP inhibits the inflammatory response induced by LPS, and thus protects mice from a lethal LPS challenge .
Several LBP-knockout mice have been generated independently, and although these mice were unable to respond to an experimental LPS challenge with cytokine release (and thus appeared to be ‘protected’), in infection models the lack of LBP was detrimental [60–63]. Thus LBP is clearly needed to fight infections caused by bacteria. It is still unclear as to whether this protective effect relies on an early detection of LPS followed by the induction of an inflammatory response, which then successfully fights bacteraemia, or whether LBP dampens the inflammatory (over-) reaction in vivo, which may be part of the systemic disease. Experiments using antibodies against LBP that later were found to not inhibit LBP activity but enhance clearance of LBP–LPS complexes, saved animals from an otherwise lethal experimental septic shock [64,65].
Other functions than LPS binding
LBP appears to be able to also bind to bacterial compounds other than LPS: It was shown early in a murine meningitis model that LBP is also a key modulator of the inflammatory response caused by the Gram-positive Streptococcus pneumoniae . While in this model it was suggested that peptidoglycan breakdown products are recognized by LBP, in other studies it has been shown that LBP can modulate the inflammatory capacity of lipopeptides, strong immunostimulatory cell wall compounds of both, Gram-positive and Gram-negative bacteria . Furthermore, Treponema and Borrelia, spirochaetes that share the double-membrane architecture with Gram-negative organisms, but clearly lack LPS, are also recognized by LBP [68–70]. Here either glycolipids or lipoproteins may be the target of LBP. Apparently, however, LBP acts as a soluble ‘pattern-recognition molecule’ and transports the bacterial ligands through the blood to present them to the appropriate TLR, inducing signal transduction. The following inflammatory response is aimed at fighting the pathogens but may also be part of the (inflammatory) disease.
Currently, there is a controversy about whether LTA (lipoteichoic acid) isolated from bacteria carries immunostimulatory properties or not. While numerous publications claimed that LTA stimulates host immune cells via TLR2, others point to the fact that minute contaminations of the LTAs with lipopeptides may have caused this effect. LBP has also been reported to modulate the effects of such an ‘LTA’ preparation and this effect may also be based on the ability to modulate the activity of lipopeptides [71,72]. Lately, a certain role for cellular LBP has been proposed. It is still unclear as to how LBP can be anchored within the cell surface; however, the presence of LBP in immune cells apparently strongly modulates the ability of these cells to respond to bacterial stimuli [73–75].
Genetic variations of LBP and disease
It has been proposed that genetic variations of the individual may also account for susceptibility and course of infectious diseases, as is well established now for cardiovascular and malignant diseases . Variations in the individual's innate immune response may determine how invading pathogens are fought, and if the inflammatory response is appropriate, too strong or prolonged. These genetic variations are probably just one cofactor, and much weaker associations as compared with monogenetic diseases should be expected. This fact makes clinical studies much harder; however, in the field of innate immunity some clear but rare associations have been found within central TLR signal transducers [77,78], and some more frequent variations within the TLR signalling system have been linked to disease susceptibility and severity [79–84]. Interestingly, the frequency of some of these TLR SNPs (single-nucleotide polymorphisms) differs largely regionally, and a selective pressure in areas of the world where fatal infectious diseases are endemic, has been proposed [85,86]. For LBP genetic variations were described early on, with first reports pointing to promoter variations potentially leading to different serum levels during the acute phase [87,88]. While no association with the outcome after myocardial infarction was found, a weak association in males was found between sepsis outcome and the presence of an LBP SNP. A recent haplotype analysis also focusing on LBP promoter variations, confirmed these results by linking a certain LBP haplotype with susceptibility to severe sepsis .
We have analysed a frequent variation of LBP in the coding domain: in vitro overexpression of the resulting LBP mutant led to a loss-of-function phenotype, also confirming a specific role for cellular LBP. Sera collected from individuals differing in their LBP genotype confirmed these results: serum from individuals with a homozygous mutation completely lacked in their ability to recognize LPS, whereas the heterozygous individuals were partly impaired. The genetic analysis of large patient cohorts revealed two separate findings (J. Eckert, K. Gürtler, S. Kaur, C. Büning, M. Kabesch, O. Kumpf and R.R. Schumann, unpublished work): the (frequent) heterozygous state seems to worsen the outcome of systemic infections, and particularly pneumonia, while the (rare) homozygous genotype is apparently associated with a chronic inflammatory disease. Thus, in the in vivo setting, intact LBP apparently is needed, and a lack of function results in an altered clinical situation. These findings will have to be repeated in larger studies, but they support the concept of LBP-mediated LPS recognition being crucial for the host.
Summary and outlook
LBP appears to be one of the better studied members of the BPI–LBP–CETP–PLTP–PLUNC family, which may be caused by the fact that it was discovered almost 25 years ago as one of the first members of this family. All the functions are not yet understood, and the three-dimensional structure still is unsolved. Within the field of endotoxin (LPS) recognition and innate immunity clearly the focus in the last 10 years has shifted towards the TLR system, and this may be a reason not too much progress in understanding LBP has been made lately. The fact that natural genetic mutations alter LBP function and change susceptibility and course of clinical diseases may help to decipher LBP function entirely. Following the complete understanding of the function, therapeutic strategies may be developed to utilize LBP to enhance defence functions and restore immune dysregulation.
Proteins with a BPI/LBP/PLUNC-Like Domain: Revisiting the Old and Characterizing the New: A Biochemical Society Focused Meeting held at New Business School, University of Nottingham, U.K., 5–7 January 2011. Organized and Edited by Colin Bingle (Sheffield, U.K.) and Sven-Ulrik Gorr (University of Minnesota School of Dentistry, Minneapolis, MN, U.S.A.).
cholesteryl ester-transfer protein
palate, lung and nasal epithelium clone
This work was supported by the Deutsche Forschungsgemeinschaft (DFG).