Lipopolysaccharide recognition: CD14, TLRs and the LPS-activation cluster

Review

TRENDS in Immunology Vol.23 No.6 June 2002

301

Lipopolysaccharide recognition: CD14, TLRs and the LPS-activation cluster Martha Triantafilou and Kathy Triantafilou Recognition of bacterial lipopolysaccharide (LPS) by the innate immune system elicits strong pro-inflammatory responses that can eventually cause a fatal sepsis syndrome in humans. LPS-mediated activation of mammalian cells is believed to involve the interaction of LPS with lipopolysaccharide-binding protein (LBP) in the serum and, subsequently with CD14. Although there is no doubt that CD14 binds LPS, CD14 is not capable of initiating a transmembrane activation signal because it is a glycosylphosphatidylinositol (GPI)-anchored protein. Accumulating evidence has suggested that LPS must interact with a transmembrane receptor(s) that is responsible for signal transduction. Integrins CD11c and/or CD18, Toll-like receptors (TLRs), as well as CD55, have been suggested to serve this function. Recently, we have revealed that a signalling complex of receptors is formed following LPS stimulation, which comprises heat-shock proteins (Hsps) 70 and 90, chemokine receptor 4 (CXCR4) and growth differentiation factor 5 (GDF5). Taking into account the discovery of the TLRs and the LPS-activation cluster, we propose a new model of LPS recognition.

Martha Triantafilou Kathy Triantafilou* Institute of Biomedical and Biomolecular Sciences, School of Biological Sciences, University of Portsmouth, King Henry Building, King Henry I Street, Portsmouth, UK PO1 2DY. *e-mail: kathy.triantafilou@ port.ac.uk

Recognition of bacterial lipopolysaccharide (LPS) by the innate immune system can lead to uncontrollable cytokine production, which can result in cardiovascular collapse and hemodynamic instability, and can eventually cause fatal sepsis syndrome in humans. Although sepsis has been increasing in incidence since the 1930s, the search for sepsis therapies has proved elusive. Gaps still exist in our knowledge of the exact events that take place on the cell membrane following LPS binding. Our current understanding of the innate immune recognition of bacterial LPS is based on the seminal discovery that LPS binds to the serum protein LPS-binding protein (LBP) [1]. LBP rapidly catalyzes the transfer of LPS to membrane-bound CD14 (mCD14) or soluble CD14 (sCD14). Although CD14 has been identified as an LPS receptor [2], it is a glycosylphosphatidylinositol (GPI)-anchored protein and thus, lacks transmembrane and intracellular domains. The mechanism(s) by which CD14 mediates a transmembrane signal have remained elusive. Several workers hypothesized that additional transmembrane receptors must act in concert with LPS–CD14 complexes to initiate the signalling process leading to LPS-induced cellular activation. This theory was strengthened by several binding studies that showed that CD14-blocking monoclonal antibodies could only partially inhibit LPS-binding, suggesting the existence of alternative receptors [3–6]. http://immunology.trends.com

Identifying the putative LPS-transducer

Over the years, work on the innate recognition of LPS has focused on identifying the putative transmembrane molecule that acts as an LPS-signal transducer using two main approaches. The first approach has been to attempt to identify molecules that bind LPS directly. Several groups utilizing different methods have identified a plethora of candidate receptors, other than CD14, ranging from 18 kDa to 96 kDa proteins [7–11]. The second approach has been to identify the Lps gene responsible for the failure of LPS to induce pro-inflammatory responses in C3H/HeJ and C57BL/10ScCr mouse strains using positional cloning. Results of long-anticipated work revealed that the Lps gene encodes the murine Toll-like receptor 4 (TLR4) [12,13]. TLR4 knockout mice demonstrated the importance of the TLR4 in LPS signalling [14]. In addition, TLR4 mutations conferred a role for this molecule in humans [15]. Furthermore, studies conducted by Shimazu et al. identified that TLR4 requires an additional molecule, MD-2, which forms a complex with the extracellular domain of TLR4, for effective LPS recognition [16]. It is now accepted that TLR4 is required for LPS signal transduction, but some questions still remain. How does TLR4 interact with CD14-LPS to transduce the signal? What are the steps involved? Are TLR4 and CD14 solely responsible for LPS binding and signal transduction? Three years after the identification of the role of TLR4 in LPS signalling, high-affinity binding of LPS to TLR4 has yet to be demonstrated. In the absence of compelling binding evidence, although we can be certain that TLR4 is involved in LPS signalling, we cannot say that LPS is a ‘ligand’ of TLR4. The challenge has now changed from simply identifying the putative LPS-‘transducer’ to determining the exact mechanism of receptor engagement. LPS-activation clusters

Recently, our group has focused on unravelling the molecular events that lead to LPS-induced signal transduction. Initial studies using fluorescence recovery after photobleaching (FRAP), a biophysical method used to determine the lateral mobility of molecules in the lipid bilayer, had revealed that LPS

1471-4906/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S1471-4906(02)02233-0

302

Review

TRENDS in Immunology Vol.23 No.6 June 2002

Table 1. Activation clustersa Cell type

Composition of activation cluster

Specificity

Refs

Monocytes/epithelial cells

HSP70, HSP90, CXCR4, GDF5

Re595 LPS

[19]

Monocytes

CD14, TLR4, CD55, CD16a, CD11b/CD18, CD32,

LPS/LTA

[20]

CD64, FcγRIIIa, CD36 and CD81 CD14, CD55, CD11b/CD18, CD32 and CD64, Ceramide

[20]

B cells

TLR4, RP105 and MD-1

FcγRIIIa, CD36 and CD47

LPS

[21]

Neutrophils/monocytes

TREM-1

LPS

[22]

Mast cells

TLR4 and MD-2

LPS

[23]

Epithelial cells

Platelet-activating factor receptor and ADAM10 TLR2 and TLR6

Staphylococcus aureus Gram-positive bacteria, yeast cell wall particle and zymosan Secreted microbial products derived from GBS Wall components of GBS

[24]

Macrophages

CD14, TLR2 and TLR6 TLRs (distinct from TLR1, -2, -4, or -6)

[25] [26] [26]

aAbbreviations:

ADAM10, a disintegrin and metalloprotease 10; GBS, Group B Streptococcus; GDF, growth and differentiation factor; HSP, heat shock protein; LPS, lipopolysaccharide; LTA; lipoteichoic acid; TLR, Toll-like receptor; TREM, triggering receptor expressed on myeloid cells.

must initially bind to CD14 and is then transferred to either an immobile receptor or a cluster of receptors [17]. The data suggest that the main function of CD14 is to catalyze the transfer of LPS from extracellular space to the membrane where it associates with a complex of receptors. This data is in good agreement with da Silva Correia et al., who have suggested that LPS must initially bind to CD14 and is then transferred to the TLR4–MD-2 complex [18]. Using affinity chromatography and peptide-mass fingerprinting we later identified a structurally heterogeneous complex of four receptors that could bind LPS. This heterogeneous receptor complex is comprised of heat shock proteins (Hsp) 70 and 90, chemokine receptor 4 (CXCR4) and growth differentiation factor 5 (GDF5) [19]. Fluorescence resonance energy transfer (FRET), another biophysical method used to study molecular interactions, revealed that LPS is indeed associated with Hsp70, Hsp90, CXCR4 and GDF5, and that all four molecules form a complex after LPS ligation. Incubation with antibodies against any of the four-identified receptors before LPS stimulation abrogated LPS-induced tumour necrosis factor (TNF)-α secretion [19]. This effect is probably because of sterical interference with other components of the receptor complex. Although TLR4 was never isolated in our biochemical experiments as an LPS-receptor, we have preliminary data suggesting that it associates with this complex of receptors and is required in order to achieve maximum response (M. Triantafilou and K. Triantafilou, unpublished). It is our belief that the mechanism of action of a couple of receptor molecules is an oversimplified model of LPS-induced activation. The existence of different supramolecular arrangements could explain how these relatively conserved individual molecules (CD14 and TLR4) have managed to cope with a rapidly changing pathogenic universe. Accumulating evidence is pointing towards the http://immunology.trends.com

direction that responses to different pathogens vary depending on cell type, composition of supramolecular activation clusters and intracellular adaptor molecules (Table 1). This model is in good agreement with Pfeiffer et al. who recently described an activation cluster composed of CD14, TLR4, CD55, CD16a, CD11b/CD18, Fcγ-receptors CD32 and CD64, Fcγ RIIIa, CD36 and CD81 after LPS or LTA stimulation [20]. Interestingly, the cluster induced by ceramide differed by the absence of CD16a, TLR4 and CD81, but presence of CD47 [20]. In addition TLR4, RP105 and MD-1 have been shown to be a part of a B-cell-specific LPS-activation cluster [21]. By contrast, triggering receptor expressed on myeloid cells (TREM)-1 has been shown to be involved in the inflammatory response caused by bacteria on neutrophils and monocytes [22]. LPS-activation clusters on mast cells seem to involve TLR4, MD-2 and the signalling molecule MyD88 [23]. Mast cells have been shown to express mRNA for TLR2 and TLR6, but not TLR5, suggesting that they have the capacity to express different TLRs and thus respond to different pathogens. Furthermore, epithelial cells seem to be utilizing platelet-activating factor receptor and ADAM10 in order to mediate responses to Staphylococcus aureus [24]. The potential for combinational diversity, in order to recognize a wide range of microbial stimuli, becomes even greater when we consider recent data suggesting the cooperation of different TLR receptors within activation clusters. TLR2 and TLR6 seem to coordinate macrophage activation by Gram-positive bacteria and the yeast cell wall particle zymosan [25]. In addition, CD14, TLR2 and TLR6 have been suggested as an activation cluster for secreted microbial products derived from Group B Streptococcus (GBS) [26]. By contrast, cell wall components of GBS are recognized by TLRs distinct from TLR1, 2, 4 or 6 [26]. It is possible that not only do different cell types utilize different ways of recognizing

Review

(a)

TRENDS in Immunology Vol.23 No.6 June 2002

TLR4

GDF5 CD55

MD-2

303

CD11/CD18

LBP LPS CD14

(b)

TLR4 MD-2

Hsp90

Hsp70

CXCR4

CD11/CD18

GDF5 CD55

CD14 LPS

CXCR4

(c)

Hsp70/90

TLR?

LPS-activation cluster GDF5 CD14

LPS

CXCR4

Hsp70/90

TLR4 CD11/CD18 MD-2 CD55

TLR?

Activation of NF-κB, MAPK, SAPK and p38 TRENDS in Immunology

Fig. 1. Hypothetical model for the immune recognition of bacteria. (a) Lipopolysaccharide-binding protein (LBP) binds and catalyses the transfer of lipopolysaccharide (LPS) to membrane-bound CD14 (mCD14). (b) Signalling molecules are recruited to the site of CD14–LPS ligation. LPS is released from CD14 in the lipid bilayer, and the intercalated LPS binds to a complex of receptors [based on our findings, chemokine receptor 4 (CXCR4), heat shock proteins (Hsps) 70 and 90, growth differentiation factor 5 (GDF5) and possibly CD55]. (c) Signal transducing molecules, such as Toll-like receptor 4 (TLR4) complexed with MD-2, Toll-like receptors (TLRs) and/or integrins CD11 or CD18 are further recruited into the activation cluster, triggering multiple signalling cascades.

LPS, but the composition of the LPS-activation clusters may change with the activation state of the cell [27]. Such a mechanism could provide an additional level of flexibility and specificity. Lessons from the acquired immune response

Thus, accumulating evidence points towards the theory that CD14 and TLR4 are not solely responsible for LPS recognition. Based on our findings, we propose a model whereby LPS initially binds to CD14. After the ligation of CD14 by LPS, different signalling molecules must be recruited to the site of ligation, where LPS is then briefly released into the lipid bilayer where it interacts with a complex of receptors, which includes Hsp70, Hsp90, CXCR4, GDF5 and TLR4 (Fig. 1). This model is in http://immunology.trends.com

good agreement with Schromm et al., whose data suggest that LPS cellular activation occurs in the plasma membrane by lateral diffusion of the intercalated LPS molecules to transmembrane proteins that then initiate signalling by steric stress [28]. Seydel et al. further suggest that the signalling protein is a stress-activated ion channel, and present data demonstrating the involvement of a high conductance, Ca2+-dependent K+ channel in LPS-induced transmembrane signalling [29]. Because CD14 is found in micro-domains, it is possible that the entire bacterial recognition system is based around the ligation of CD14 by bacterial products and the formation of a signalling complex of receptors at the site of CD14–LPS ligation within membrane micro-domains. By analogy to the immunological synapse, it is also possible that two micro-domains or ‘zones’ of activation exist: a periphery and a core region. Once CD14 releases LPS, it might stay in the periphery, whereas all the other signalling molecules are gathered in the ‘core’ region of the micro-domain. Because different receptors are recruited to the site of ligation forming an activation cluster, multiple signalling cascades are triggered. Thus, the existence of a receptor cluster explains the variety of signalling cascades that are triggered by

304

Review

TRENDS in Immunology Vol.23 No.6 June 2002

LPS. LPS has been shown to be able to activate the NF-kB, ERK1/2 and SAPK/JNK pathways. It is possible that different cell types and different supramolecular complexes utilize different intracellular pathways. For example, Watters et al. have demonstrated ERK activation in a murine macrophage line, but also ERK-independent signalling in murine microglia [30]. The existence of a receptor cluster responsible for bacterial recognition could explain why the search for anti-sepsis therapies has been so elusive. Therapeutic interventions targeting single components of this complex could be futile. In the future, therapeutic approaches targeting multiple receptors should be considered. Conclusions and future direction

We propose that the innate recognition of bacteria involves the dynamic association of multiple receptors within one or more micro-domains. The transient association of different receptors within the activation cluster could give rise to the recognition and/or discrimination of different ligands. Given the diverse range of receptors involved in these activation clusters, the potential for different combinations of receptors to recognize a wide range of microbial stimuli is enormous. This References 1 Schumann, R.R. et al. (1990) Structure and function of lipopolysaccharide-binding protein. Science 249, 1429–1431 2 Wright, S.D. et al. (1990) CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding-protein. Science 249, 1431–1433 3 Lynn, W.A. et al. (1993) Neither CD14 nor serum is absolutely necessary for activation of mononuclear phagocytes by bacterial lipopolysaccharide. Infect. Immun. 61, 4456–4461 4 Blondin, C. et al. (1997) Lipopolysaccharide complexed with soluble CD14 binds to normal human monocytes. Eur. J. Immunol. 27, 3303–3309 5 Troelstra, A. et al. (1997) Saturable CD14-dependent binding of fluorescein-labelled lipopolysaccharide to human monocytes. Infect. Immun. 65, 2272–2275 6 Triantafilou, M. et al. (2000) Rough and smooth forms of fluorescein-labelled bacterial endotoxin exhibit CD14/LBP dependent and independent binding that is influenced by endotoxin concentration. Eur. J. Biochem. 267, 2218–2226 7 Lei, M.G. et al. (1988) Specific endotoxin lipopolysaccharide binding protein on murine splenocytes. J. Immunol. 141, 996–1005 8 Dziarski, R. (1991) Peptidoglycan and lipopolysaccharide bind to the same binding site on lymphocytes. J. Biol. Chem. 266, 4719–4725 9 Kirkland, T.N. et al. (1990) Identification of lipopolysaccharide binding proteins in 70Z/3 cells by photoaffinity crosslinking. J. Biol. Chem. 265, 9520–9525 10 El-Samalouti, V.T. et al. (1997) Detection of lipopolysaccharide (LPS)-binding membrane proteins by immuno-coprecipitation with LPS and anti-LPS antibodies. Eur. J. Biochem. 250, 418–424

http://immunology.trends.com

potential becomes even greater when you consider that at least some of the ten known TLR family members can heterodimerize (e.g. TLR2/1 and TLR6/1), giving rise to an even broader ligand specificity [25]. It is our view that, based on the lack of compelling binding data, TLRs might well be downstream of the initial step or steps of LPS recognition. It could be that TLR4 is recruited later on and is required to achieve maximum response (Fig. 1c). Future work should focus on unravelling the exact molecular events that lead to LPS-induced cellular activation and attempt to answer the most vexing questions that have arisen since the discovery of the LPS-activation cluster. How do all these membrane proteins fit together in the LPS ‘sensing apparatus’? In what time frame and with what stochiometry do they associate? Do these interactions occur within micro-domains? And finally, what is the exact multi-molecular choreography of receptors involved in bacterial recognition? Only with progress towards the latter can we hope to understand better the innate recognition of bacteria and try to target therapeutically not just single receptors, but the plethora of molecules that seem to play a crucial role in the development of septic shock.

11 Triantafilou, K. et al. (2001) Interactions of bacterial lipopolysaccharide and peptidoglycan with a 70kDa and an 80kDa protein on the cell surface of CD14 positive and CD14 negative cells. Hum. Immunol. 62, 50–63 12 Poltorac, A. et al. (1998) Defective LPS signaling in C3H/Hej and C57BL/10ScCr mice: mutations in TLR4 gene. Science 28, 2085–2088 13 Qureshi, S.T. et al. (1998) Endotoxin-tolerant mice have mutations in toll-like receptor 4 (TLR4). J. Exp. Med. 189, 615–625 14 Hoshino, K. et al. (1999) Cutting edge: toll-like receptor 4 (TLR4) deficient mice are hyporesponsive to lipopolysaccharide: evidence for TLR4 as the LPS gene product. J. Immunol. 162, 3749–3752 15 Schwartz, D.A. (2001) The role of TLR4 in endotoxin responsiveness in humans. J. Endotoxin Res. 7, 389–393 16 Shimazu, R. et al. (1999) MD-2, a molecule that confers lipopolysaccharide responsiveness on Toll-like receptor 4. J. Exp. Med. 189, 1777–1782 17 Triantafilou, K. et al. (2001) Fluorescence recovery after photobleaching reveals that lipopolysaccharide rapidly transfers from CD14 to heat shock proteins 70 and 90 on the cell membrane. J. Cell Sci. 114, 2535–2545 18 da Silva Correia, J. et al. (2001) Lipopolysaccharide is in close proximity to each of the proteins in its membrane receptor complex. Transfer from CD14 to TLR4 and MD-2. J. Biol. Chem. 276, 21129–21135 19 Triantafilou, K. et al. (2001) A CD14independent LPS receptor cluster. Nat. Immun. 4, 338–345 20 Pfeiffer, A. et al. (2001) Lipopolysaccharide and ceramide docking to CD14 provokes ligandspecific receptor clustering in rafts. Eur. J. Immunol. 31, 3153–3164

21 Ogata, H. et al. (2000) The Toll-like receptor protein RP105 regulates lipopolysaccharide signalling in B cells. J. Exp. Med. 192, 23–29 22 Bouchon, A. et al. (2001) TREM-1 amplifies inflammation and is crucial mediator of septic shock. Nature 410, 1103–1105 23 McCurdy, J.D. et al. (2001) Toll-like receptor 4-mediated activation of murine mast cells. J. Leukocyte. Biol. 70, 977–984 24 Lemjabbar, H. and Basbaum, C. (2002) Plateletactivating factor receptor and ADAM10 mediate responses to Staphylococcus aureus in epithelial cells. Nat. Med. 8, 41–46 25 Ozinsky, A. et al. (2000) The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between Toll-like receptors. Proc. Natl. Acad. Sci. U. S. A. 97, 13766–13771 26 Henneke, P. et al. (2001) Novel engagement of CD14 and multiple Toll-like receptors by Group B Streptococci. J. Immunol. 167, 7069–7076 27 Zarember, K.A. and Godowski, P.J. (2002) Tissue expression of human Toll-like receptors and differential regulation of Toll-like receptor mRNAs in leukocytes in response to microbes, their products and cytokines. J. Immunol. 168, 554–561 28 Schromm, A. et al. (2000) Biological activities of lipopolysaccharides are determined by the shape of their lipid A portion. Eur. J. Biochem. 267, 2008–2014 29 Seydel, U. et al. (2001) A K+ channel is involved in LPS signaling. J. Endotoxin Res. 7, 243–247 30 Watters, J.J. et al. A differential role for the mitogen-activated protein kinases in LPS signaling: the MEK/ERK pathway is not essential for nitric oxide and interleukin 1β production. J. Biol. Chem. 277, 9077–9087