FASEB J. Mp Biomedicals
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.06-7579com.
 (The FASEB Journal. 2007;21:3231-3239.)
© 2007 FASEB
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow All Versions of this Article:
fj.06-7579comv1
21/12/3231    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wong, K.-F.
Right arrow Articles by Fan, S.-T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wong, K.-F.
Right arrow Articles by Fan, S.-T.

Characterization of two novel LPS-binding sites in leukocyte integrin ßA domain

Kwong-Fai Wong*,1, John M. Luk*,{ddagger},1,2, R. Holland Cheng{dagger}, Lloyd B. Klickstein{ddagger} and Sheung-Tat Fan*

* Department of Surgery, University of Hong Kong, Pokfulam, Hong Kong;

{dagger} Department of Molecular and Cellular Biology, University of California, Davis, California, USA;

{ddagger} Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA

2Correspondence: Department of Surgery, The University of Hong Kong, Jockey Club Clinical Research Center, 21 Sassoon Rd., Pokfulam, Hong Kong. E-mail: jmluk{at}hkucc.hku.hk


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Lipopolysaccharide (LPS), a bacterial endotoxin, triggers deleterious systemic inflammatory responses when released into blood circulation, causing organ dysfunction and death. In response to LPS stimulation, CD14 and toll-like receptor (TLR)-4 elicit inflammatory signaling cascades. Although leukocyte integrins (CD11b/CD18 and CD11c/CD18) were reported to bind LPS and induce NF-{kappa}B translocation, the evidence on such epitope location remains elusive. The present study aims to delineate the LPS-binding sites on the integrin CD18 antigen and to design peptide(s) as potential prophylactic and/or therapeutic agents to modulate LPS effects in activated Jurkat cells. Epitope mapping analysis using a series of CD18 truncated variants revealed two putative LPS-binding sites within the ßA region (216–248 and 266–318 a.a.), which were further confirmed by point mutation studies. Inhibition assay demonstrated that the CD18-ßA266–318 peptide could block LPS binding in a dose-dependent manner. Our data also indicated that treatment with the CD18-peptide modulated TNF-{alpha} mRNA transcription via the NF-{kappa}B signaling pathway in LPS-activated Jurkat cells. In conclusion, we have identified two novel LPS-binding sites located at the CD18 ßA domain of leukocyte integrin, and the integrin peptide ßA266–318 is shown to inhibit LPS binding and subsequent inflammatory events, having therapeutic implications to cure Gram-negative endotoxemia.— Wong, K-F, Luk, J. M., Cheng, R. H., Klickstein, L. B., and Fan, S-T. Characterization of two novel LPS-binding sites in leukocyte integrin ßA domain.


Key Words: lipopolysaccharide • endotoxin-neutralizing peptide • NF-{kappa}B • sepsis • prophylaxis • TNF-{alpha}


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
LIPOPOLYSACCHARIDE (LPS) IS A MAJOR constituent of the outer membrane of Gram-negative bacteria (1) . Having been known as endotoxin, LPS can initiate systemic inflammatory response by activating a variety of monocytic cells and other leukocytes, stimulating the release of proinflammatory cytokines and up-regulating the expression of endothelial adhesion molecules. Systemic release of LPS into blood circulation causes lethal endotoxemia, a clinical sequel following major surgical operations, deep wound infections, and trauma among those critically ill patients in intensive care units (2) .

Elucidation of signaling pathway central to LPS-mediated proinflammatory responses may allow us to design therapeutic targets for curing endotoxemia; and in this context CD14 and toll-like receptor 4 (TLR4) are identified as the two key LPS receptors on monocytic cells (3 , 4) . Due to the hydrophobicity of LPS, it requires the presence of LPS-binding protein (LBP) as carrier in serum to promote the dissociation of LPS micelles into monomers and accelerate the binding of these monomers to CD14 antigen on leukocyte surface (5) . Thereafter, CD14-LPS complex would interact with the TLR4 to transduce MyD88-dependent inflammatory signaling cascade that involves multiple transcriptional regulators, including nuclear factor-kappa B (NF-{kappa}B) and p38 mitogen activated protein kinase (MAPK) for the synthesis of tumor necrosis factor (TNF)-{alpha} and interleukin (IL)-1 (6 , 7) . However, previous studies have shown that monoclonal antibodies targeting CD14 antigen yielded limited success in clinical trials, and no significant protection of the septic shock patients was observed (8) . In vitro studies also indicated that CD14 antibody could only partially inhibit LPS binding (9) and was unable to eradicate TNF-{alpha} production in monocytes (10) . Thus, it is believed that alternative transduction mediators other than CD14 may be responsible for the pathogenesis of Gram-negative endotoxemia, particularly in the CD14-deficient lineages.

Recent studies have identified leukocyte ß2 integrin as a candidate for LPS receptor apart from their well-known adhesive functions. Leukocyte ß2 integrins (CD11/CD18) are heterodimers in which a common CD18 antigen (ß subunit) pairing with at least four distinct {alpha} subunits, viz. CD11a (of LFA-1), CD11b (of Mac-1), CD11c (of p150/95), and CD11d. Transfection of LFA-1 and p150/95 genes into the CHO fibroblast cell line rendered cellular responsiveness to LPS stimulation (11 , 12) . In addition, Mac-1 has been found to act in concert with CD14 and TLR4 to induce LPS-mediated cytokine expressions (13) . Because the sequences of integrin {alpha} chains are rather heterogeneous among different members, it is thought that the LPS-binding site is likely residing in the common ß subunit. Indeed, the engagement of CD18 in LPS interactions was noted by using an anti-CD18 antibody to block LPS and E. coli binding to macrophages (14) . Furthermore, the N-terminal region of CD18 has been suggested as the binding domain of leukotoxin in bovine leukocyte integrins (15) .

Despite these initial findings, the exact epitope location for LPS recognition along the CD18 ß chain remains unknown. Within the CD18 antigen, the primary sequence of ßA (or the "I-like") domain is relatively conserved and involved in ligand recognition including intercellular adhesion molecule-1 (ICAM-1), C3bi, and neutrophil inhibitory factor (16) . Given these information, we were prompted to investigate any LPS-binding activities of the CD18 antigen and to delineate the epitope location in the ßA domain. Clinically, it would be of great interest to examine whether the derived CD18 peptide sequence that is responsible for LPS interaction could be exploited as a novel therapeutic LPS antagonist to intervene the endotoxin-mediated tissue damages.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Bacterial lipopolysaccharide (LPS)
Purified LPS from Salmonella minnesota (Re595 and R7) and Salmonella typhimurium (TV119 and SL1181) were purchased from Sigma (St. Louis, MO, USA), while LPS from Bordetella pertussis and Rhodobacter sphaeroides were obtained from LIST Biological Laboratories (Campbell, CA, USA). Biotinylation of Salmonella minnesota Re595 LPS was performed as described (17) . LPSs were purified by chromatography and phenol-extraction procedures and protein contamination of the preparations was less than 1% as determined by the manufacturers. All solution and buffer used in this study were prepared in endotoxin-free water.

Preparation of ßA domain, truncation, and point-mutation variants
To facilitate epitope mapping of putative LPS-binding sites resident in leukocyte ß2 integrin, three overlapping truncated fragments (~120 residues in length): a.a.104–222 (N-terminal), 167–285 (central piece), and 223–341 (C-terminal) of the ßA domain were constructed and tested for the corresponding LPS-binding activities by direct ELISA assay (Fig. 1 ). Afterward, the N-terminal piece was found with no detectable reactivity, and thus, fine mapping with three shorter fragments (~60 a.a. residues) was designed from the sequence 167–341. The LPS reactivity was detected in peptide sequences of 226–284 and 285–341, respectively. When the ßA domain was subjected to BLAST search for homologous sequences (as described below), the 216–248 and 266–318 sequences were shown to share similarity with the certain LPS-binding motifs: bactericidal/permeability increasing protein (BPIP), LBP, limulus anti-LPS factor (LALF), as well as cationic antibacterial protein of 18kDa (CAP18), respectively. These results were in alignment with the above two described truncated fragments 226–284 and 285–341. In the following studies, we essentially focused on the ßA peptide sequences 216–248 and 266–318 for their reactivities with LPS.


Figure 1
View larger version (18K):
[in this window]
[in a new window]

  		Figure 1. Delineation of LPS-binding motif. Left panel: Schematic presentation of the epitope mapping analysis for the LPS-binding sites along the ßA domain of CD18 antigen. Truncated fragments were cloned, expressed, and assayed for LPS-binding activities. Right panel: ELISA OD reading (405 nm) of each truncated fragments, recombinant peptides, denatured protein, and no antigen controls. The data shown were representative from three independent experiments, mean ± SD (triplicates). The initial epitope mapping analysis identified the ßA sequence from 226 to 341 responsible for the Salmonella minnesota Re595 LPS binding activities.

We have recently reported the expression of recombinant ßA domain with different fusion partners. The human CD18 cDNA construct was a kind gift from Dr. Timothy A. Springer (Harvard Medical School, Boston, MA, USA). For the present study, truncated ßA domain fragments were amplified by PCR and cloned into pET43.1-B expression vector (Novagen, San Diego, CA, USA) within the SacI and HindIII restriction sites, which contained a short HSV tag at the C-terminus to recombinant protein. Full-length of wild-type ßA domain and mutant ßA-K317A were also constructed without the C-terminal HSV tag, and subcloned into pET43.1-C vector (Novagen), followed by restriction digestion with NotI and XhoI whose recognition sites were enclosing the HSV tag coding sequence. Point mutations of the ßA domain were constructed using the PCR-based overlap-extension approach as described (18) . Briefly, in the first-round of PCR, two fragments with overlapping region enclosing the desired mutations were amplified using the mutagenic oligonucleotides (Invitrogen, Carlsbad, CA, USA) as listed in Table 1 . Annealing of the two fragments was then performed as follows: (a) 7 cycles of denaturation at 94°C for 1 min and annealing at 60°C for 4 min. Resulting mutants were amplified by another (b) 25 cycles of denaturation at 94°C for 1 min, annealing at 55°C for 30 s, and extension at 72°C for 1 min. Subsequent cloning of the ßA mutants followed the same procedures as described above for the wild-type. All cDNA constructs were verified by DNA sequencing analysis to confirm the inframe mutations introduced using the Genetic Analyzer 3100 (Applied Biosystems Inc., Forest Hill, CA, USA).


View this table:
[in this window]
[in a new window]

  		Table 1. Mutagenic oligonucleotides for constructions of ßA domain mutants

Recombinant ßA domain and its variants were produced in Escherichia coli strain BL21 (DE3) (Novagen) in LB medium, after induction by 0.5 mM of isopropyl-beta-D-thiogalactopyranoside (IPTG) (Amersham Biosciences, Piscataway, NJ, USA). Periplasmic soluble fractions were harvested by sonication in lysis buffer (50 mM Tris-HCl, pH 8.0; 300 mM NaCl; and 10 mM imidazole), and recombinant proteins carrying N-terminal poly-(His)6 tag were purified by IMAC affinity chromatography using the Ni-NTA agarose column (Qiagen, Valencia, CA, USA). After running 3 column volumes with washing buffer (50 mM Tris-HCl, pH 8.0; 0.3 M NaCl; and 20 mM imidazole), protein was eluted with 200 mM imidazole, dialyzed in sterile PBS, and examined by 10% SDS-PAGE and Western blot (19) . The final concentration of purified protein was measured by the standard Bradford assay (Bio-Rad, Hercules, CA, USA).

Cell cultures and protein lysates
Jurkat cells were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 100 mg/ml penicillin G and 50 µg/ml streptomycin at 37°C and 5% CO2. About 1 x 106 cells per ml were challenged with Salmonella minnesota Re595 LPS (2 µg/ml), followed by treatments with sterile PBS or recombinant ßA peptides (10 µM). After 4-h incubation, cells were harvested for mRNA and protein analyses. All experiments were performed in triplicate and repeated at least twice. For preparation of cell lysate for binding study, cells were first harvested by centrifugation and resuspended in lysis buffer (20 mM Tris-HCl, pH 7.4; 0.15 M NaCl; 2 mM EDTA; 10% glycerol; 1% Nonidet P-40; 1 mM PMSF). Cell lysate was then centrifuged at 1500 g for 5 min, and supernatant obtained was collected for experiments.

LPS binding assays
LPS binding activity of recombinant ßA domain toward lipopolysaccharides from Salmonella minnesota, Salmonella typhimurium, Bordetella pertussis, and Rhodobacter sphaeroides were first determined by ELISA as described previously (17 , 20) , and Salmonella minnesota Re595 LPS was chosen for the epitope mapping study. Briefly, Re595 LPS was coated at 5 µg/ml in bicarbonate buffer (0.1 M, pH 9.6) on 96-well flat-bottom microtiter plate (Corning, Corning, NY, USA) at 4°C overnight. After washing and blocking with 1% BSA in PBS, purified recombinant ßA proteins (wild-type, truncated variants, or point mutants) with HSV tag were incubated in serial dilutions for 2 h at 37°C. Bound proteins were detected by horseradish peroxidase (HRP)-conjugated anti-HSV antibody (Novagen), followed by substrate development of 2'-azino-bis (3-ethylbenzthiazoline-6-sulfonic acid (ABTS) (Invitrogen) and measured at 405 nm by ELISA microplate reader (Molecular Devices, Sunnyvale, CA, USA). To confirm the binding specificity, inhibition experiments were performed using the full-length ßA wild-type and ßA-K317A mutant proteins without the HSV tag as inhibitors in varying concentrations from 0 to 10 µM. Recombinant HSV-tagged ßA domain, ßA216–248 and ßA266–318 peptides at fixed concentration of 10 µM were allowed to react with the Salmonella minnesota Re595 LPS antigen (5 µg/ml) precoated onto microtiter plate in the presence of varying concentrations (0 to 10 µM) of ßA wild-type or ßA-K317A mutant protein. Sterile PBS buffer was used as baseline control, and the following incubation and detection procedures were similar as described above. For blocking studies of LPS binding to Jurkat cells, protein lysates (10 µg/ml) in the bicarbonate buffer was coated to the plate at 4°C overnight. After saturation with 1% BSA in PBS, the plate was added with biotinylated Salmonella minnesota Re595 LPS (5 µg/ml) in the presence of i) sterile PBS, ii) bovine serum albumin (BSA) (Amersham Biosciences), or iii) ßA peptides (serial concentrations from 0 to 8 µM) at 37°C for 60 min. Bound biotinylated LPS was detected by addition of HRP-streptavidin conjugate (1:5000 dilution) and followed by ABTS substrate development as described above.

Real-time quantitative PCR
Total RNA was extracted from Jurkat cells by Trizol reagent (Invitrogen) and DNaseI digestion (Ambion, Austin, TX, USA), following manufacturers’ instruction. Reverse transcriptions were carried out using 1 µg RNA and TaqMan Reverse Transcriptase kit (Applied Biosystems). Quantitative PCR amplification of the TNF-{alpha} gene was carried out using the Power SYBR Green PCR Master Mix (Applied Biosystems) in ABI PRISM 7700 sequence detector system (Applied Biosystems). After normalization with ß-actin internal house-keeping control, TNF-{alpha} transcript levels were compared among different treatment groups as previously mentioned (21) .

Immunoblotting of NF-{kappa}B in nuclear extract
Nuclear extract of LPS-stimulated Jurkat cells was prepared as reported in details (22) . Briefly, nuclei were harvested from cell lysate by centrifugation at 1500 g for 5 min, and nuclear contents were released by resuspending pellets in high salt buffer (20 mM HEPES, pH 7.9; 25% (v/v) glycerol; 0.42 M NaCl; 1.5 mM MgCl2; 0.2 mM EDTA; 0.5 mM PMSF; and 0.5 mM DTT). The NF-{kappa}B level in nuclear extract was determined by Western blot using mouse anti-NF-{kappa}B (p65) monoclonal antibody (1:2500 dilution; Invitrogen). Nuclear factor levels were normalized by histone content that was detected by mouse anti-histone monoclonal antibody (1:3000 dilution; Chemicon, Temecula, CA, USA).

Bioinformatic analysis and molecular modeling
Protein sequences of bactericidal/permeability increasing protein (BPIP) (23) , LBP (24) , limulus anti-LPS factor (LALF) (25) , and cationic antibacterial protein of 18 kDa (CAP18) (26) were available from NCBI database (http://www.ncbi.nlm.nih.gov/Database/index.html) and sequence alignments with CD18 ßA domain or integrin peptides were performed by ClustalW program available from the European Bioinformatics Institute (http://www.ebi.ac.uk/clustalw) (27) . Secondary structure prediction was calculated by JPred (http://www.compbio.dundee.ac.uk/~www-jpred) (28) , and the helical wheel diagram was constructed by BioEdit (http://www.mbio.ncsu.edu/BioEdit/bioedit.html) (29) . In silico modeling of ßA domain was generated by SWISS-MODEL server (http://swissmodel.expasy.org) (30) , and the resulting model was fine-tuned or modified by RasMol (http://www.umass.edu/microbio/rasmol/).

Statistic analysis
All statistical analyses were calculated using the SPSS software for Window version 13 (SPSS, Inc., Chicago, IL, USA). Results were presented as mean ± SD. Statistical significance was evaluated by Student’s t test, and P < 0.05 values were considered significant.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Characterization of the LPS-binding sites along CD18 ßA domain
To facilitate the epitope mapping of unknown LPS binding sites resident at the ßA domain, we cloned a series of truncated ßA fragments and assessed their binding activities to Salmonella minnesota Re595 LPS antigen by ELISA. First, three major fragments (N-terminal: a.a. 104–222; central piece: a.a 167–285; C-terminal: a.a 223–341) spanning through the entire ßA domain were constructed and subjected to LPS binding assay. As shown in Fig. 1 , the data indicated that the sequence from 167 to 341 residues retained ~80–90% of LPS-binding activity when compared with the full-length ßA domain. On the contrary, the N-terminal fragment (a.a: 104–222) showed weak or no detectable activity. Thus, we focused on the region (167–341) and constructed three shorter truncated fragments (each ~60 residues) covering this region. It was found that LPS-binding activity was retained in fragments corresponding to residues 226–284 and 285–341, whereas the N-terminal fragment (a.a 167–225) gave weak or undetectable LPS-binding activity.

Acquisition of these regions prompted us to further delineate the location of the LPS-binding site(s) through bioinformatic approach. The integrin ßA domain and peptides (226–284 and 285–341) were subjected to BLAST search for homologous LPS recognition sites. Using ClustalW, we retrieved two regions ßA216–248 and ßA266–318 (as exemplified by the open rectangles in Fig. 2 ), which aligned themselves to the published endotoxin-binding motifs. For example, ßA216–248 (216DAMMQVAACPEEIGWRNVTRLLVFATDDGFHFA248) was aligned to three distinct proteins including BPIP, LBP, and LALF with similarity of ~24%; whereas ßA266–318 (266LEDNLYKRSNEFDYPSVGQLAHKLAENNIQPIFAVTSRMVKTYEKLTEIIPKS318) was aligned to CAP18 with 43% similarity. To determine whether these individual binding sites could sustain LPS binding activities, recombinant integrin peptides were generated and both peptides were found to bind LPS in a dose-dependent manner comparable to that of the full-length ßA domain (Fig. 3 A). Notably, the ßA216–248 yielded relatively weaker activity. Binding specificity of these two recombinant integrin peptides, ßA216–248 and ßA266–318 was further confirmed by inhibition with full-length wild-type ßA domain without the HSV tag, but not by the mutant ßA-K317A protein (Fig. 3B ).


Figure 2
View larger version (17K):
[in this window]
[in a new window]

  		Figure 2. Protein sequence alignment analysis of the putative LPS-binding region. The ßA domain was aligned against the known endotoxin-binding motifs available from published literatures using the ClustalW alignment program. Identical residues are highlighted in black boxes; whereas residues with similar physicochemical property are gray shadowed. Two regions namely (A) ßA216–248 and (B) ßA266–318 were retrieved. Similarity between the ßA216–248 and the BPI, LBP, and LALF was ~24% (A), whereas ßA266–318 showed similarity of 43% with the CAP18 (B). Open rectangles represented the integrin peptides designed for further binding and functional studies.


Figure 3
View larger version (22K):
[in this window]
[in a new window]

  		Figure 3. Characterization of LPS-binding motifs. A) Direct ELISA assay: Evaluation of the LPS-binding activities of recombinant peptides representing the ßA216–248 and ßA266–318 regions. Varying concentrations of integrin peptides (with HSV tag) were added to the Salmonella minnesota Re595 LPS antigen (5 µg/ml) coated onto 96-well microtiter plates, and bound peptides were detected by the HRP-conjugated anti-HSV antibody as described in the Materials and Methods. Color intensity was measured by ELISA reader at 405 nm. Heat-denatured recombinant wild-type ßA domain protein was used as control of the experiment. The data shown were representative from three independent experiments, mean ± SD (triplicates). B) Binding of recombinant HSV-tagged full-length ßA domain, ßA216–248 and ßA266–318 peptides of fixed concentration (10 µM) to immobilized LPS in the presence of varying concentrations of full-length ßA domain (wild-type) and ßA-K317A (mutant control). Binding of ßA domain and the ßA peptides were presented as percentage binding when compared to the PBS (buffer) control. For (C, D), effect of point-mutants introduced into various Arginine (R->A) and Lysine (K->A) sites at the ßA domain was assayed for LPS-binding activities by ELISA as described above. Heat-denatured ßA protein served as control. For simplicity, binding curves of point mutations along the two LPS-binding motifs (C) ßA216–248 and (D) ßA266–318 were plotted separately. Data shown were representative of three independent experiments, and error bars represented the SD of triplicates.

Given the importance of cationic interaction in LPS accommodation, point mutational analysis was conducted in order to evaluate the importance of the cationic residues along the two integrin peptides (ßA216–248: Arg231 and Arg235; ßA266–318: Arg273, Arg303, Lys306, Lys310 and Lys317) where potentially harbored the LPS-binding sites. A pilot study showed that mutations did not affect the production of solubleprotein; and mutants matched the size as expected from the primary sequences (Supplemental Fig. 1A ). Substitutions of Arg303 and lysine resides (Lys306, Lys310, and Lys317) in the ßA266–318 peptide resulted in severe loss of LPS binding activity. Likewise, point mutations of the ßA216–248 also affected the normal LPS binding to certain extent; however, the cationic residues along ßA216–248 (Fig. 3C ) were more tolerant of substitutions when compared with their counterparts along ßA266–318 (Fig. 3D ).

Inhibition of ßA peptides on LPS binding to Jurkat cells
Previous studies have demonstrated that binding of LPS to its cellular receptors could be blocked by peptides derived from other known endotoxin-binding domains (26) . To test whether this is true in the case of integrin peptides, we examined the inhibitory effects of ßA216–248 and ßA266–318 peptides for their ability to intervene the binding of biotinylated Salmonella minnesota Re595 LPS to Jurkat lysates using our previously established method (17) . As shown in Fig. 4 , the ßA266–318 peptide, but not the ßA216–248, markedly suppressed the LPS binding to Jurkat cell, and the inhibition was dose-dependent. Incubation with 8 µM of ßA266–318 peptide yielded 50% inhibition. By contrast, ßA216–248 peptide did not illustrate to the same extent. Inhibition of biological activities is a complex system, and probably the epitope location may have influence on the response. In this case, ßA216–248 peptide binds to LPS but is not critical enough to block binding of LPS to Jurkat cells whereas ßA266–318 peptide is sufficient per se to abrogate LPS interaction.


Figure 4
View larger version (15K):
[in this window]
[in a new window]

  		Figure 4. Blockage of LPS binding to Jurkat cells by integrin ßA peptides. Biotinylated Salmonella minnesota Re595 LPS was allowed to incubate with coated Jurkat cell lysates in the presence of BSA (Control), ßA216–248 peptide, ßA266–318 peptide, or PBS alone. Integrin peptides were diluted at varying concentration and incubated with the biotinlylated LPS solution. Bound LPS was determined by HRP-streptavidin conjugate followed by substrate development. Optical density was measured at 405 nm. Binding of LPS was presented as percentage binding when compared to the PBS (no inhibitor) control. The experiment was done in triplicate, and data shown were mean ± SD. All experiments were repeated at least thrice. The integrin ßA266–318 peptide demonstrated significant inhibition at ≥6 µM. *P < 0.05.

Modulation of TNF-{alpha} expression by ßA peptides
As the most potent proinflammatory cytokine, the TNF-{alpha} level has been found up-regulated on LPS challenge, and this up-regulation is mediated by NF-{kappa}B signaling pathway (31) . In accordance with the previous findings, the TNF-{alpha} transcript level of Jurkat cells was demonstrated to escalate by near 3-fold after challenge with Salmonella minnesota Re595 LPS (Fig. 5 A). Nuclear accumulation of NF-{kappa}B was also noted to increase accordingly (Fig. 5B ). With the inhibitory effect of ßA peptides on LPS binding to Jurkat as demonstrated above, we next examined the potential immunomodulatory effect of LPS-stimulated TNF-{alpha} by the ßA peptides. Jurkat cells were allowed to incubate with the LPS and ßA peptides. In the presence of ßA266–318 peptide, the up-regulation of TNF-{alpha} expression mediated by LPS challenge in Jurkat was found significantly inhibited (P=0.001), and the TNF-{alpha} level appeared to be near physiological level (Fig. 5A , right panel). As determined by Western blotting (Fig. 5B ), immunoreactivity of NF-{kappa}B in nuclear extract of cell preparation was markedly suppressed after treatment with the ßA266–318 peptide. Again, under the same condition, equivalent amount of ßA216–248 peptide manifested no significant effects on the TNF-{alpha} transcript level, nor the NF-{kappa}B level in nuclear extract.


Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		Figure 5. Modulation of ßA216–248 and ßA266–318 peptides on LPS-induced TNF-{alpha} expression in Jurkat cells. Jurkat cells were challenged with LPS (2 µg/ml) and administrated with ßA216–248 peptide, ßA266–318 peptide, or PBS alone. A) Effect of ßA peptides on TNF-{alpha} transcription as determined by real-time quantitative-PCR. Data shown were fold changes in target gene levels after receiving treatments as compared with the PBS control. Experiments were done in duplicates and repeated at least thrice. B) Immunoreactivity of NF-{kappa}B in nuclear fraction. Western blot was conducted using the anti-p65 monoclonal antibody to probe against the nuclear extracts from different treatment groups of LPS-activated Jurkat cells. Antihistone antibody was included as loading control.

Identification of putative LPS-interacting structural elements
Sequence analysis and in silico modeling of ßA domain revealed structural elements that may favor interaction with LPS. Given the importance of helical conformation involved in endotoxin binding, we first located the helical conformation along the two putative LPS-binding sites by secondary prediction using the JPred algorithm. The retrieved helical conformations were located at 219MQVAA223 (of ßA216–248), 282VGQLAHKLAE291 (of ßA266–318), and 302SRMVKTYEKLTEI314 (of ßA266–318). Further detailed analysis of these helical conformations by BioEdit revealed the 302SRMVKTYEKLTEI314 of ßA266–318 formed an amphipathic helical structure (Fig. 6 A). In silico modeling of ßA domain was also performed to portray the spatial arrangement of the two LPS-binding sites (Fig. 6B ). It was noted that the ßA266–318 was more solvent-exposed than the ßA216–248, probably rendering this binding site more accessible to accommodate the LPS binding. Importantly, it was also appreciated that the carboxyl end of ßA266–318 was depicted to fold into an amphipathic helix.


Figure 6
View larger version (29K):
[in this window]
[in a new window]

  		Figure 6. Protein sequence analysis and in silico modeling of integrin ßA domain. A) Secondary structure of ßA domain was predicted by JPred and as shown in which ‘H’ and ‘E’ represent {alpha}-helix and ß-strand, respectively. Arrows and asterisks represent the start and the end of the two LPS-binding sites, respectively. Drawing of helical wheel diagram of boxed sequence 302SRMVKTYEKLTEI314 was constructed by BioEdit. A helix with an amphipathic helical conformation, which is characterized by the presence of both hydrophilic and hydrophobic sectors was illustrated. In silico modeling of ßA domain. Predicted tertiary structure was constructed by SWISS-MODEL, and the molecular models were visualized by RasMol. B) Molecular model of the integrin ßA domain portraying the spatial arrangements of i: ßA216–248 binding site; ii: ßA266–318 binding site; and iii: both binding sites together. The starting residues of each binding sites are also indicated (green). C) Different views of the same molecular model of ßA266–318 peptide with the side chains of Arg303, Lys306, Lys310, and Lys317 highlighted: (i) front view and (ii) top view of the amphipathic helix.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In the present study, we successfully identified two novel LPS-binding sites within the CD18 ßA domain of leukocyte ß2 integrins and also demonstrated preliminary findings to implicate the recombinant ßA peptides as therapeutic agents potentially to intervene endotoxin-mediated proinflammatory response by activated leukocytes during Gram-negative sepsis. Early studies on LPS recognition by innate immunity suggested that ß2 integrins served as a receptor for bacteria and bacterial endotoxin (14 , 32) . However, no further biochemical or molecular information was available. Our present findings provided detailed studies about the LPS recognition sites by epitope mapping analysis. Based on the binding assay using a series of truncated domain variants and bioinformatics sequence-alignment analysis, we have determined the two putative LPS-binding sites located at residues 216–248 and 266–318 along the ßA domain. To our knowledge, this is the first report on characterization of LPS binding to the integrin ßA domain. It was different from that reported the residues 1–291 of bovine CD18 constituted the binding domain for leukotoxin secreted by Mannheimia (Pasteurella) hemolytica (15) . Furthermore, sequence alignment and in silico molecular modeling analysis revealed that these two LPS sites exhibited analogy with the existing endotoxin-binding motifs in coordination of physiochemical-like residues. It is believed that the integrin ßA peptides are likely binding to the conserved region of the endotoxin of different Gram-negative species (supplementary Figure S1B).

Of great importance is how these two sites may interact with the LPS molecule. Indeed, elucidation of protein-LPS interaction has been challenging due to the heterogeneity and aggregative property of LPS (33) . A partial understanding that illustrates the involvement of three-dimensional geometric arrangement of cationic side chain has evolved from a previous structural study, in which an iron uptake protein (FhuA) from E. coli. membrane was cocrystallized with LPS (34) . Thus, we examined the differential LPS binding activities of various ßA mutants that would hopefully depict how ß2 integrins accommodate endotoxin. Our mutational analysis indicated that Arg303 and lysine residues (Lys306, Lys310, and Lys317) along ßA266–318 were required for LPS binding. Further evidence was hinted by the location of these residues within the amphipathic {alpha}-helical conformation (Fig. 6C ). This tertiary structure was resembled to CAP18 [Protein Data Bank (PDB) ID code 1LYP] and reported to favor interaction with the negatively-charged LPS (35) . Contrary to the ßA266–318, cationic residues along the ßA216–248 were found more tolerant of substitutions and no distinct structural elements favoring charge-charge interactions were identified. Collectively, our observation anticipated that the mechanisms by which the two binding sites accommodated LPS would be different; ßA266–318 bound LPS through cationic interaction while ßA216–248 probably utilized other interactions like hydrophobic one that has not been predicted herein yet.

Cellular binding to and subsequent receptor activation of immune cells by endotoxin or bacterial LPS have drawn immense interests in fields of innate immunity. Despite its heterogeneity from Gram-negative bacterial species, the LPS molecules conform to a common structural principle composed of highly polymorphic polysaccharide chains and a well-conserved lipid A moiety (the endotoxic center). Both receptor binding and subsequent activation are believed to require the participation of lipid A. Early experiments utilizing a variety of chemically modified LPSs suggested that the hydrophilic portion was of importance in mediating the receptor binding while the unique conformation of hydrophobic portion (acyl groups) was of central significance in activating target cells (36) . This paradigm of separation of LPS binding and activation favors the development of peptides deriving from the endotoxin-binding motifs, which can potentially neutralize the biological actions of LPS. In the present study, we identified two putative LPS-binding sites within the ßA domain, and the ßA266–318 peptide, but not the ßA216–248, demonstrated the inhibitory effect on LPS binding and alleviated TNF-{alpha} transcription in LPS-activated Jurkat cells. It is thus suggested that the two peptides likely bind to different sites of the LPS molecule and that the binding of ßA266–318 peptide could either directly antagonize receptor activation or indirectly alter the lipid A conformation insomuch that the biological activity is affected. Nevertheless, the molecular details or exact mechanistic of the ßA266–318 peptide and what physical property conferring to the anti-LPS effect remain to be determined. Previous study on cathelicidin family of antibacterial peptides proposed that the LPS-neutralizing activity of a peptide was correlated with its cationicity (26) . Along this line, we compared the isoelectric values of defensins with the ßA266–318 peptide and found that despite {alpha}-defensin (human neutrophil peptide-1) and ß-defensins (hBD-1 and hBD-2) showed higher pI values (8.28, 8.55, and 9.25, respectively) than the ßA266–318 peptide (apparent pI=6.78), they have not ever been found to suppress the LPS action on RAW264.7 cells. Thus, this information may enlighten our understanding on the action of endotoxin-neutralizing peptide that the anti-LPS effect may not be associated with cationicity, but rather, probably attributed to the distinct structural features (in this case, the amphipathic {alpha}-helical conformation in ßA266–318 peptide).

During antibiotic therapy against Gram-negative bacteria, cell lysis of bacteria cause release of LPS. Considered as the primary target of LPS, monocytic cells are stimulated to release proinflammatory mediators during antibiotic therapy against severe bacteremia. Among those cytokines, TNF-{alpha} is the key player to cause sepsis development secondary to endotoxemia (37) . It has been generally accepted that scavenging of the circulating LPS is beneficial to amelioration of endotoxin-mediated tissue injury. Because of its LPS blockage and neutralizing properties, integrin peptide derived from ßA266–318, alone or in combinations with antibody therapies targeting to LPS (38 , 39) , holds promise to ameliorate endotoxemia and reduce the mortality of critically ill patients due to Gram-negative sepsis. Nevertheless, the efficacy and safety of using integrin ßA peptide or its derivatives in curing endotoxemia required further evaluation in animal models.


   ACKNOWLEDGMENTS
 
The authors thank the critical reading and comments of Dr. Nikki Lee, and the study was supported by the Research Grants Council of Hong Kong (HKU 7320/02M).


   FOOTNOTES
 
1 These authors contributed equally to this work.

Received for publication January 27, 2007. Accepted for publication April 19, 2007.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Osborn, M. J., Rosen, S. M., Rothfield, L., Zeleznick, L. D., Horecker, B. L. (1964) Lipopolysaccharide of the Gram-negative cell wall. Science 145,783-789[Abstract/Free Full Text]
  2. Rice, T. W., Bernard, G. R. (2005) Therapeutic intervention and targets for sepsis. Annu. Rev. Med. 56,225-248[CrossRef][Medline]
  3. Wright, S. D., Ramos, R. A., Tobias, P. S., Ulevitch, R. J., Mathison, J. C. (1990) CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 249,1431-1433[Abstract/Free Full Text]
  4. Poltorak, A., He, X., Smirnova, I., Liu, M. Y., Van Huffel, C., Du, X., Birdwell, D., Alejos, E., Silva, M., Galanos, C., Freudenberg, M., Ricciardi-Castagnoli, P., Layton, B., Beutler, B. (1998) Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282,2085-2088[Abstract/Free Full Text]
  5. Hailman, E., Lichenstein, H. S., Wurfel, M. M., Miller, D. S., Johnson, D. A., Kelley, M., Busse, L. A., Zukowski, M. M., Wright, S. D. (1994) Lipopolysaccharide (LPS)-binding protein accelerates the binding of LPS to CD14. J. Exp. Med. 179,269-277[Abstract/Free Full Text]
  6. Yamamoto, M., Yamazaki, S., Uematsu, S., Sato, S., Hemmi, H., Hoshino, K., Kaisho, T., Kuwata, H., Takeuchi, O., Takeshige, K., Saitoh, T., Yamaoka, S., Yamamoto, N., Yamamoto, S., Muta, T., Takeda, K., Akira, S. (2004) Regulation of Toll/IL-1-receptor-mediated gene expression by the inducible nuclear protein IkappaBzeta. Nature 430,218-222[CrossRef][Medline]
  7. Ozinsky, A., Underhill, D. M., Fontenot, J. D., Hajjar, A. M., Smith, K. D., Wilson, C. B., Schroeder, L., Aderem, A. (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[Abstract/Free Full Text]
  8. Reinhart, K., Gluck, T., Ligtenberg, J., Tschaikowsky, K., Bruining, A., Bakker, J., Opal, S., Moldawer, L. L., Axtelle, T., Turner, T., Souza, S., Pribble, J. (2004) CD14 receptor occupancy in severe sepsis: results of a phase I clinical trial with a recombinant chimeric CD14 monoclonal antibody (IC14). Crit. Care Med. 32,1100-1108[CrossRef][Medline]
  9. Lynn, W. A., Liu, Y., Golenbock, D. T. (1993) Neither CD14 nor serum is absolutely necessary for activation of mononuclear phagocytes by bacterial lipopolysaccharide. Infect. Immun. 61,4452-4461[Abstract/Free Full Text]
  10. Gessani, S., Testa, U., Varano, B., Di Marzio, P., Borghi, P., Conti, L., Barberi, T., Tritarelli, E., Martucci, R., Seripa, D., et al (1993) Enhanced production of LPS-induced cytokines during differentiation of human monocytes to macrophages. Role of LPS receptors. J. Immunol. 151,3758-3766[Abstract]
  11. Flaherty, S. F., Golenbock, D. T., Milham, F. H., Ingalls, R. R. (1997) CD11/CD18 leukocyte integrins: new signaling receptors for bacterial endotoxin. J. Surg. Res. 73,85-89[CrossRef][Medline]
  12. Ingalls, R. R., Golenbock, D. T. (1995) CD11c/CD18, a transmembrane signaling receptor for lipopolysaccharide. J. Exp. Med. 181,1473-1479[Abstract/Free Full Text]
  13. Perera, P. Y., Mayadas, T. N., Takeuchi, O., Akira, S., Zaks-Zilberman, M., Goyert, S. M., Vogel, S. N. (2001) CD11b/CD18 acts in concert with CD14 and Toll-like receptor (TLR) 4 to elicit full lipopolysaccharide and taxol-inducible gene expression. J. Immunol. 166,574-581[Abstract/Free Full Text]
  14. Wright, S. D., Jong, M. T. (1986) Adhesion-promoting receptors on human macrophages recognize Escherichia coli by binding to lipopolysaccharide. J. Exp. Med. 164,1876-1888[Abstract/Free Full Text]
  15. Gopinath, R. S., Ambagala, T. C., Deshpande, M. S., Donis, R. O., Srikumaran, S. (2005) Mannheimia (Pasteurella) haemolytica leukotoxin binding domain lies within amino acids 1 to 291 of bovine CD18. Infect. Immun. 73,6179-6182[Abstract/Free Full Text]
  16. Zhang, L., Plow, E. F. (1996) Overlapping, but not identical, sites are involved in the recognition of C3bi, neutrophil inhibitory factor, and adhesive ligands by the alphaMbeta2 integrin. J. Biol. Chem. 271,18211-18216[Abstract/Free Full Text]
  17. Luk, J. M., Kumar, A., Tsang, R., Staunton, D. (1995) Biotinylated lipopolysaccharide binds to endotoxin receptor in endothelial and monocytic cells. Anal. Biochem. 232,217-224[CrossRef][Medline]
  18. Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K., Pease, L. R. (1989) Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77,51-59[CrossRef][Medline]
  19. Lee, N. P., Tsang, S., Cheng, R. H., Luk, J. M. (2006) Increased solubility of integrin betaA domain using maltose-binding protein as a fusion tag. Protein Pept. Lett. 13,431-435[CrossRef][Medline]
  20. Luk, J. M., Lind, S. M., Tsang, R. S., Lindberg, A. A. (1991) Epitope mapping of four monoclonal antibodies recognizing the hexose core domain of Salmonella lipopolysaccharide. J. Biol. Chem. 266,23215-23225[Abstract/Free Full Text]
  21. Lo, C. Y., Lam, K. Y., Leung, P. P., Luk, J. M. (2005) High prevalence of cyclooxygenase 2 expression in papillary thyroid carcinoma. Eur. J. Endocrinol. 152,545-550[Abstract/Free Full Text]
  22. Dignam, J. D., Lebovitz, R. M., Roeder, R. G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 11,1475-1489[Abstract/Free Full Text]
  23. Little, R. G., Kelner, D. N., Lim, E., Burke, D. J., Conlon, P. J. (1994) Functional domains of recombinant bactericidal/permeability increasing protein (rBPI23). J. Biol. Chem. 269,1865-1872[Abstract/Free Full Text]
  24. Battafaraono, R. J., Dahlberg, P. S., Ratz, C. A., Johnston, J. W., Gray, B. H., Haseman, J. R., Mayo, K. H., Dunn, D. L. (1995) Peptide derivatives of three distinct lipopolysaccharide binding proteins inhibit lipopolysaccharide-induced tumor necrosis factor-alpha secretion in vitro. Surgery 118,318-324[CrossRef][Medline]
  25. Hoess, A., Watson, S., Siber, G. R., Liddington, R. (1993) Crystal structure of an endotoxin-neutralizing protein from the horseshoe crab, Limulus anti-LPS factor, at 1.5 A resolution. EMBO J. 12,3351-3356[Medline]
  26. Nagaoka, I., Hirota, S., Niyonsaba, F., Hirata, M., Adachi, Y., Tamura, H., Heumann, D. (2001) Cathelicidin family of antibacterial peptides CAP18 and CAP11 inhibit the expression of TNF-alpha by blocking the binding of LPS to CD14(+) cells. J. Immunol. 167,3329-3338[Abstract/Free Full Text]
  27. Thompson, J. D., Higgins, D. G., Gibson, T. J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22,4673-4680[Abstract/Free Full Text]
  28. Cuff, J. A., Clamp, M. E., Siddiqui, A. S., Finlay, M., Barton, G. J. (1998) JPred: a consensus secondary structure prediction server. Bioinformatics 14,892-893[Abstract/Free Full Text]
  29. Hall, T. A. (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids. Symp. Ser. 41,95-98
  30. Schwede, T., Kopp, J., Guex, N., Peitsch, M. C. (2003) SWISS-MODEL: An automated protein homology-modeling server. Nucleic Acids Res. 31,3381-3385[Abstract/Free Full Text]
  31. Heumann, D., Roger, T. (2002) Initial responses to endotoxins and Gram-negative bacteria. Clin. Chim. Acta 323,59-72[CrossRef][Medline]
  32. Moore, K. J., Andersson, L. P., Ingalls, R. R., Monks, B. G., Li, R., Arnaout, M. A., Golenbock, D. T., Freeman, M. W. (2000) Divergent response to LPS and bacteria in CD14-deficient murine macrophages. J. Immunol. 165,4272-4280[Abstract/Free Full Text]
  33. Brandenburg, K., Mayer, H., Koch, M. H., Weckesser, J., Rietschel, E. T., Seydel, U. (1993) Influence of the supramolecular structure of free lipid A on its biological activity. Eur. J. Biochem. 218,555-563[Medline]
  34. Ferguson, A. D., Hofmann, E., Coulton, J. W., Diederichs, K., Welte, W. (1998) Siderophore-mediated iron transport: crystal structure of FhuA with bound lipopolysaccharide. Science 282,2215-2220[Abstract/Free Full Text]
  35. Andra, J., Gutsmann, T., Garidel, P., Brandenburg, K. (2006) Mechanisms of endotoxin neutralization by synthetic cationic compounds. J. Endotoxin. Res. 12,261-277[Medline]
  36. Rietschel, E. T., Kirikae, T., Schade, F. U., Mamat, U., Schmidt, G., Loppnow, H., Ulmer, A. J., Zahringer, U., Seydel, U., Di Padova, F., et al (1994) Bacterial endotoxin: molecular relationships of structure to activity and function. FASEB J. 8,217-225[Abstract]
  37. Hack, C. E., Aarden, L. A., Thijs, L. G. (1997) Role of cytokines in sepsis. Adv. Immunol. 66,101-195[Medline]
  38. Bone, R. C., Balk, R. A., Fein, A. M., Perl, T. M., Wenzel, R. P., Reines, H. D., Quenzer, R. W., Iberti, T. J., Macintyre, N., Schein, R. M. (1995) A second large controlled clinical study of E5, a monoclonal antibody to endotoxin: results of a prospective, multicenter, randomized, controlled trial. The E5 Sepsis Study Group. Crit. Care Med. 23,994-1006[CrossRef][Medline]
  39. Warren, H. S., Amato, S. F., Fitting, C., Black, K. M., Loiselle, P. M., Pasternack, M. S., Cavaillon, J. M. (1993) Assessment of ability of murine and human anti-lipid A monoclonal antibodies to bind and neutralize lipopolysaccharide. J. Exp. Med. 177,89-97[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
M. J. Scott and T. R. Billiar
{beta}2-Integrin-induced p38 MAPK Activation Is a Key Mediator in the CD14/TLR4/MD2-dependent Uptake of Lipopolysaccharide by Hepatocytes
J. Biol. Chem., October 24, 2008; 283(43): 29433 - 29446.
[Abstract] [Full Text] [PDF]

Home page
 Infect. Immun.Home page
J. F. Charles, M. B. Humphrey, X. Zhao, E. Quarles, M. C. Nakamura, A. Aderem, W. E. Seaman, and K. D. Smith
The Innate Immune Response to Salmonella enterica Serovar Typhimurium by Macrophages Is Dependent on TREM2-DAP12
Infect. Immun., June 1, 2008; 76(6): 2439 - 2447.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow All Versions of this Article:
fj.06-7579comv1
21/12/3231    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wong, K.-F.
Right arrow Articles by Fan, S.-T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wong, K.-F.
Right arrow Articles by Fan, S.-T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS