| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* National Research Laboratory of Defense Proteins, College of Pharmacy and Research Institute for Drug Development, Pusan National University, Busan, Korea;
Research Center Borstel, Leibniz-Center for Medicine and Biosciences, Medical and Biochemical Microbiology, Borstel, Germany; and
Department of Comparative Physiology, Evolutionary Biology Center, Uppsala University, Uppsala, Sweden
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-D-glucan (1, 2). Pattern recognition receptors expressed on the cell surface or in the cell, including the family of Drosophila Toll, human TLRs, or nucleotide oligomerization domains, can activate signaling pathways that induce antimicrobial effector responses and inflammation upon recognition of different PAMPs (3, 4). Also, the typical secreted pattern recognition proteins in human plasma, such as mannose binding lectin and ficolin, can recognize PAMPs and promote the elimination of microbes through the activation of phagocytosis or the lectin pathway of the complement system (5). Several soluble pattern recognition proteins, such as PGN recognition protein (PGRP) (6), 1,3--D-glucan recognition protein (GRP) (7), and LPS and 1,3--D-glucan recognition protein (8) have been identified in invertebrates. They have been shown to be involved in the activation of the prophenoloxidase (proPO) activating system that in many aspects is similar to the mammalian lectin pathway of the complement system (1). Even though several biological functions of soluble pattern recognition proteins have been determined, their molecular recognition mechanism still remains to be clarified.
The proPO cascade is an important non-self-recognition system present in most invertebrates. It is an enzyme cascade containing several serine proteases and their inhibitors, and terminates with the zymogen proPO. Microbial polysaccharides LPS, PGN, and 1,3--D-glucan first react with pattern recognition proteins, which then induce activation of several serine proteases within the proPO system. Determining the molecular mechanism, by which pattern recognition molecules distinguish non-self from self and transduce signals that stimulate defense responses, is a key for understanding the ways in which innate immune systems are regulated.
Previously, we reported (9, 10, 11) the structures and functions of three soluble pattern recognition proteins, Holotrichia PGRPs and Tenebrio GRP, and Tenebrio PGRP from two large beetles, Holotrichia diomphalia and Tenebrio molitor larvae. Holotrichia PGRPs recognized both bacterial PGN and fungal -1,3-D-glucan (9), but Tenebrio GRP specifically recognized only fungal 1,3--D-glucan (10). Tenebrio PGRP recognized both the Lys-type and diaminopimelic acid-type PGN (11). These soluble pattern recognition proteins are involved in 1,3--D-glucan- or PGN-dependent proPO activation. Recently, we reported (12) the crystal structure of a clip-domain serine protease and functional roles of the clip domains during proPO activation. We also reported (13) that apolipoprotein III from the greater wax moth, Galleria mellonella, binds to fungal conidia and -1,3-D-glucan, and, therefore, may act as a pattern recognition molecule for multiple microbial and parasitic invaders. However, the molecular structure of the LPS recognition protein (LRP) and the mechanism by which LPS induces activation of the proPO system in invertebrates are not fully understood.
Accumulating evidences suggest that LPS, a major cell wall component of Gram-negative bacteria, is known for their ability to initiate humoral and cellular immune responses and is important for maturation of the immune system of vertebrates or invertebrates. It was reported (14) that LPS is involved in many aspects of the innate immune system and proteins showing specific affinity to LPS are thought to be important in the innate immune reactions. Therefore, coleopteran insects could have a LRP to be able to prevent invasion of pathogenic Gram-negative bacteria by triggering innate immune reactions such as phagocytosis, encapsulation, and activation of the proPO system. In preliminary experiments, we observed an Escherichia coli-agglutinating activity when crude cell-free Holotrichia plasma was incubated with E. coli. This agglutination phenomenon was inhibited by the addition of LPS in vitro, suggesting that an unidentified novel LRP(s) may exist in the plasma. One reason that made it difficult to purify LRP from this insect hemolymph is that the activation of the proPO system is easily triggered and then melanin synthesis is induced, which results in nonspecific protein-protein cross-linking (15). This latter process prevents attempts to purify novel LRP(s) by column chromatography.
To identify and investigate the biological role of a novel LRP from the plasma of H. diomphalia larvae, we prepared a solution with a nonactivated proPO system by treating plasma with an irreversible serine protease inhibitor, diisopropyl fluorophosphate (DFP), to prevent activation of the proPO-activating enzyme(s). By using DFP-treated plasma, we isolated a novel 40-kDa protein to homogeneity, cloned its cDNA, and examined its function in vivo and in vitro. This 40-kDa protein contained six repeats of an epidermal growth factor (EGF)-like domain that recognized the inner core oligosaccharides of LPS. In addition, it also enhanced the clearance of E. coli from insect hemolymph. In this study, we report, for the first time, a novel function in innate immunity for a 40-kDa protein containing EGF-like domains.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
The purified LPS and lipid A were obtained from Borstel Research Center (Hamburg, Germany), according to previously published methods (16). Briefly, smooth (S) -type LPS of E. coli serotypes O4, O6, O12, O15, and O18 and of Salmonella enterica serovar abortus equi and serovar Friedenau were prepared by the phenol-water procedure (17) and then purified by repeated ultracentrifugation and converted into sodium salt form by electrodialysis (18). Rough LPS of E. coli R1 (F470), R2 (F576), R3 (F653), R4 (strain 2513), and K-12 as well as LPS from core-deficient strains of S. enterica serovar minnesota chemotypes Ra (strain R60), Rb2 (strain R345), Rc (strain R5), Rd1 (strain R7), and Rd2 (strain R4) and E. coli Re (strain F515) were obtained by the phenol-chloroform-petroleum-ether method (19). Free lipid A in its bisphosphorylated form was obtained from E. coli Re mutant F515 by treating phenol-chloroform-petroleum-ether-extracted LPS with acetate buffer (20). The purified LPS and lipid A were solubilized in water and stored in aliquots at 1 mg/ml at 4°C. The lyophilized LPS, E. coli O111:B4, and E. coli O55:B5 were obtained from Sigma-Aldrich.
Collection of hemolymph and preparation of DFP-treated plasma solution
Methods for rearing the insects and collection of hemolymph were as previously described (21). To prevent the activation of the proPO-activating enzyme(s), 750 µl of 2 mM DFP was added to 150 ml of plasma solution (10 µM DFP as a final concentration) and the mixture was incubated for 2 h at 4°C, and then plasma was dialyzed for 12 h at 4°C against buffer A (50 mM Tris-HCl buffer, pH 7.5) and is referred to as DFP-treated plasma. An assay of phenoloxidase (PO) was conducted according to our previously published method (22).
Purification of E. coli-agglutinating protein and cyanogen bromide (CNBr) treatment
The DFP-treated plasma (150 ml, 1 g of proteins) was applied onto an agarose gel column (3.0 x 15 cm, Bio-gel A-5m gel; Bio-Rad) equilibrated with buffer A, and washed extensively with the same buffer until the absorbance at 280 nm was 0. The bound proteins were eluted with buffer A containing 0.1, 0.3, and 0.5 M NaCl, sequentially, and examined for E. coli-agglutinating activity. The active fractions were pooled, concentrated with ultrafiltration (YM10; Amicon), and then dialyzed against 12 liters of buffer B (50 mM Tris (pH 7.5) containing 0.15 M NaCl) for 12 h at 4°C. The dialyzed sample (4.4 mg of proteins) was loaded onto a Sephacryl S-200 gel filtration column (2 x 130 cm) previously equilibrated with buffer B. The gel filtration was performed in buffer B and fractions showing E. coli agglutinating activity were analyzed by SDS-PAGE under reducing conditions and then active fractions were pooled. Ab against the purified 40 kDa was raised by injecting 10 µg of the purified protein into a male albino rabbit with CFA, and 14 days later, two booster injections with the same amount of protein were done. For purification of the E. coli- binding motif of LRP, this protein was treated by CNBr to generate several peptides basically as described in Ref. 23 . Briefly, lyophilized purified LRP (50 µg) dissolved in 80% formic acid was adjusted to 10 mg/ml CNBr in 80% formic acid and incubated under N2 overnight at room temperature (RT). Cleavage was terminated by the addition of 10 volumes of H2O, and proteins were lyophilized, dissolved in 5% acetic acid, and purified by C18 reverse phase-HPLC. After the examination of E. coli-agglutinating activity, the identities of the obtained peaks were verified by the determination of their amino-terminal sequences.
cDNA cloning and nucleotide sequencing of LRP
To determine the internal amino acid sequences of the purified 40-kDa protein, the purified protein (25 µg) was reduced, alkylated, and digested with 2 µg of lysylendopeptidase (WAKO) at 37°C for 12 h. The resulting peptides were separated through HPLC on a reverse phase C18 column (Waters) and applied to an Applied Biosystems Procise Automated Gas Phase Sequencer. For determination of the amino-terminal sequence, the purified protein (1 µg), transferred to a polyvinylidene difluoride membrane, was directly applied to the protein sequencer. A cDNA library from H. diomphalia larvae was constructed with a previously described method (22), using a Zap-cDNA synthesis kit (Stratagene). Among five internal sequences of the purified 40-kDa protein, DNA oligonucleotide corresponding to GQCYGNFC was synthesized as follows: 5'-GGICA(A/G)TG(T/C)TA(T/C)GGIAA(T/C)TT(T/C)TG-3' and it was labeled with [
-32P]ATP. For the initial screening, 
120,000 recombinants of the cDNA library were used. We obtained 10 hybridization-positive clones and analyzed two plasmids containing five internal and amino-terminal sequences. The deduced amino acid sequence of the 40 kDa was compared with the protein sequence database of the National Center for Biotechnology Information.
Expression and purification of rLRP The cDNA encoding the mature LRP was ligated into a vector called pFASTBACtevFc. The pFASTBACtevFc is a modified from pFASTBAC-HTc (Invitrogen Life Technologies) to express secreted proteins, fused with the Fc domain of human IgG at the C terminus. The Fc domain can be cleaved from the resulting protein by thrombin or TEV protease due to thrombin and tobacco etch virus (TEV) protease recognition sites are inserted before the Fc domain gene in the vector. For large production of the rLRP, Spodoptera frugiperda 9 (SF-9; Invitrogen Life Technologies) cells were grown at 27°C in 1 liter of Grace’s medium (Invitrogen Life Technologies) supplemented with 20% Sf-900II serum-free medium and 0.1% Pluronic F-68 reagent. The cells with a density of 2 x 106 cells/ml were infected at a multiplicity of infection of 1–5, and then were incubated for 36 h. The medium containing the rLRP was harvested by centrifugation at 2300 x g for 10 min. The supernatant was incubated with 2 ml of protein A-Sepharose (Amersham Biosciences) at 4°C for 1 h, which was then washed with buffer C (20 mM Tris-HCl containing 200 mM NaCl, pH 8.0). The rLRP-bound resins were pooled and incubated with 20 U of thrombin (100 U/ml) overnight at 4°C, then was eluted with buffer C. The rLRP was concentrated to 0.5 mg/ml for additional experiments. The identity of the purified rLRP was confirmed by immunoblotting analysis using LRP Ab.
Measurement of E. coli-agglutinating activity and competition experiments with the purified LPS and lipid A
The agglutination activity was examined by the incubation of 3 µl of bacteria (3 x 106 cells of E. coli Top10F or S. aureus ATCC 6538 or fungi (Candida albicans TIMM1776) with 1.5 µl of LRP (1.5 µg) for 30 min at RT, and then agglutinating activity of LRP was studied under the microscope. To test whether E. coli-agglutinating activity of the native or rLRP was inhibited by various purified S-type or rough-type LPS, LPS (10 µg/10 µl) was preincubated with the native or rLRP (3 µg/3 µl) with 5 µl of PBS for 20 min at RT, and then E. coli (3 x 106 cells) were added to the mixture. After 10 min, the E. coli-agglutinating activity was estimated by the microscope.
Scanning electron microscopy
Bacterial and fungal cells were prepared for electron microscopy essentially by following the protocol used to determine the agglutinating activity above, except that sterile round glass coverslips (13 mm) were placed in the cell culture wells. The cells were fixed in 2.5% glutaraldehyde (24) and then washed for 3.5 min in PBS (pH 7.0), postfixed in 1% osmium tetroxide for 2 h, and washed again three times with PBS. Dehydration was conducted in an ascending ethanol series to 95% overnight. Two further dehydration steps in 99% ethanol for 20 min were conducted before drying in critical point dryer and gold coating. Images were captured with a Philips XL30 electron microscope.
Confocal microscope
Experiments were performed with a laser scanning confocal microscope (LSM 510 META; Zeiss) coupled to an inverted microscope (Axiovert 200M; Zeiss). Cells were excited with a 633-nm wavelength HeNe laser, and images were acquired using a >560-nm long-pass filter. Fluorescence of MitoTracker Red/MeOH (Molecular Probes) was excited at 560 nm and emitted at 600 nm. To load fluorescence dye onto E. coli, bacterial cells were washed two times with PBS and resuspended with PBS to a final concentration of 1 x 109cells/ml. E. coli (5 x 108 cells) were incubated with MitoTracker Red CMXRos (1 µM) for 20 min and washed two times with PBS. The agglutination activity was examined by incubation of 3 µl of bacteria (3 x 106 cells) and rLRP (3 µg) for 30 min at RT, and then rLRP-induced E. coli-agglutinating activity was examined.
Preparation of LRP-deficient plasma
To examine the function of LRP, we prepared LRP-deficient plasma using an affinity-purified LRP Ab. The purified LRP (500 µg) was coupled to 2 ml of CNBr-Sepharose (0.5 g resin) following the manufacturer’s protocol. Five milliliters of rabbit serum (80 mg/ml) containing polyclonal Abs against LRP was diluted with 10x volume of buffer A and applied to an LRP-coupled column (2 ml) pre-equilibrated with buffer A. After washing, bound Abs were eluted with 100 mM glycine (pH 2.5) and immediately neutralized with 1 M Tris-HCl until pH was 7.0. To prepare the LRP-deficient plasma solution, a solution containing an affinity-purified LRP Ab (130 µg/50 µl) was mixed with protein A-resin (30 µl of 50% suspended protein A-Sepharose CL-4B) for 30 min at 4°C, and then DFP-treated Holotrichia plasma solution (1 mg/130 µl) was added to the above solution and further incubated for 30 min at 4°C. The mixture solution was centrifuged at 2500 x g for 5 min at 4°C. The supernatant was used as the LRP-deficient plasma. The absence of a 40-kDa LRP in this plasma was confirmed by immunoblotting analysis.
E. coli clearance experiments on Holotrichia larvae
In vivo bacterial clearance was tested as described by Koizumi et al. (25). Briefly, four batches of newly cultured E. coli were prepared. First, E. coli (2 x 108 cells) were incubated in 500 µl of DFP-treated Holotrichia plasma containing 8 mg of protein at 4°C for 1 h to obtain E. coli coated with the LRP from the crude plasma (crude LRP E. coli). Second, the same numbers of E. coli cells were incubated in 500 µl of PBS containing 50 µg of the purified LRP at 4°C for 1 h to obtain E. coli coated with the purified LRP (purified LRP-E. coli). Third, the same numbers of cells were incubated in 500 µl of PBS as a positive control (control E. coli). Fourth, the same numbers of bacteria cells were coincubated with an affinity-purified LRP Ab (50 µg) for 30 min at 37°C (Ab E. coli). After incubation, the mixtures were centrifuged at 2500 x g for 15 min at 4°C, then supernatants were removed, and the precipitated E. coli were washed three times with PBS. The recovered cells from four batches were recounted by hemocytometer, and then bacteria (2 x 106 cells) were resuspended in 50 µl of PBS (4 x 104 cells/1 µl). Ten microliters of each batch sample was injected to one larva. Treated larvae were kept at 25°C, and 500 µl of hemolymph from each larva was collected at different time periods (30 min, 1 h, 3 h, and 6 h). The collected hemolymphs were centrifuged at 2500 x g for 10 min at 4°C, and the precipitated hemocytes and E. coli cells were resuspended in 500 µl of PBS; 100 µl of each sample was plated immediately on a Luria-Bertani plate and incubated overnight at 37°C, and the number of viable bacteria was counted.
Preparation of mouse bone marrow mast cells (BMMC) and neutralization of LPS by LRP
Mouse BMMC were prepared as described previously (26). More than 95% of the viable cells were confirmed to be mast cells, as assessed by staining with acidic toluidine blue and the measurement of c-kit expression by FACS. Neutralization of LPS was measured as the loss of its ability to secrete IL-6 production of BMMC as described previously (27). Briefly, after stimulation of mouse BMMC with LPS (300 ng/ml) alone or coincubation with LPS (300 ng/ml) and LRP (150 ng/ml) for 30 min at 37°C, IL-6 levels were measured using a commercially available mouse IL-6 ELISA kit (R&D Systems).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
To identify LRP in the insect hemocoel, we first examined E. coli-agglutinating activity by using H. diomphalia crude hemolymph. As shown in Fig. 1B, crude hemolymph exhibited E. coli-agglutinating activity, indicating that Holotrichia hemolymph may contain an unidentified LRP. When we examined the agglutinating activity by using cell-free plasma or hemocyte lysate, only plasma exhibited E. coli- agglutinating activity (Fig. 1C), but not hemocyte lysate (Fig. 1D), suggesting that an unidentified E. coli-agglutinating molecule(s) is present in the plasma.
|
When we tried to purify the E. coli agglutination molecule from the crude Holotrichia plasma, it was impossible to carry out column chromatographies because of a strong PO activity of plasma and melanin pigments formed rapidly through the activation of the proPO system. Therefore, we concluded that the Holotrichia proPO system must be maintained in a nonactivated form for a successful purification of the LRP(s). Previously, we reported (21) that Holotrichia proPO-activating factors involved in proPO activation were strongly inhibited by DFP. As expected, the DFP-treated plasma (•, Fig. 2A) did not show any PO activity compared with the crude plasma (
, Fig. 2A), indicating that Holotrichia proPO activating factors were inactivated by DFP treatment. To obtain a novel LRP from Holotrichia plasma, we used an agarose gel. When the bound proteins were eluted with 0.1, 0.3, and 0.5 M NaCl, sequentially, fractions eluted with 0.3 and 0.5 M NaCl showed E. coli-agglutinating activity, but not the 0.1 M NaCl fraction (data not shown). The active fractions were analyzed by SDS-PAGE under reducing conditions (Fig. 2B, lane 4). Because a 40-kDa major protein and several minor bands in the active fractions were obtained, we conducted a Sephacryl S-200 gel filtration column for further purification. The elution profile of Sephacryl S-200 is shown in Fig. 2C and each peak was analyzed by SDS-PAGE under reducing conditions (Fig. 2D, lanes 1-3). E. coli-agglutinating activity was shown in the second peak. A 40-kDa protein showing E. coli-agglutinating activity was found to be mostly enriched compared with the crude hemolymph (Fig. 2D, lane 2). Approximately 1.5 mg of the 40-kDa protein could be purified reproducibly from 1 g of the Holotrichia plasma protein. We determined five internal and amino-terminal sequences. These were IVAHTPFHXTPDXPSGXGL, IXXSGYK, XISVXPGRXLNGQXYGNFXN, GAGFGPDGK, SGYALNS, and TTNPTAPRAYRPRINTTAGV (NH2-terminal sequence), in which X indicates an unidentified residue. When these five amino acid sequences of the purified 40-kDa protein were compared with the National Center for Biotechnology Information data, two partial sequences showed homology with those of tenascin and Notch that contain EGF-like domains.
|
The nucleotide sequence of the longest insert and the deduced amino acid sequence of the 40-kDa protein encoded in the open reading frame of this clone have been submitted to the DNA Data Bank in Japan as accession number AB201466.4 The open reading frame encodes a protein consisting of 317 aa residues. All five internal and amino-terminal sequences derived from the 40-kDa protein are found within the complete sequence, indicating that the cDNA codes for the 40-kDa protein (data not shown). The deduced amino acid sequence of the 40-kDa protein did not show any significant homology with other proteins so far reported, but it contains six repeats of an EGF-like domain (Fig. 3A). The EGF-like domain is found in various molecules including Drosophila Notch (28), Drosophila Delta (29), human fibrillin (30), chicken tenascin (31), and human EGF (Ref. 32 and Fig. 3B). It has been reported that human EGF has a profound effect on the differentiation of specific cells in vivo and is a potent mitogenic factor for a variety of cultured cells of both ectodermal and mesodermal origin (32). However, no report has shown that a protein containing EGF-like domains has E. coli-agglutinating activity.
|
To ascertain the biological function of the 40-kDa protein for E. coli agglutination activity, we purified native LRP and rLRP to homogeneity from the DFP-treated plasma solution and from a baculovirus expression system, respectively (Fig. 2D, lanes 2 and 4). To confirm E. coli-agglutinating activity of the purified 40-kDa protein, we examined E. coli-agglutinating activity with the purified LRP by using scanning electron microscopy. Fig. 4 shows intact E. coli (A), agglutinated E. coli by the purified LRP (B and C). However, this E. coli agglutination was not induced in the presence of LPS from E. coli O111 (Fig. 4D). Under the same conditions, S. aureus and C. albicans cells were not agglutinated by the purified LRP (Fig. 4, E and F, respectively), suggesting that the purified LRP may function as an LPS recognition protein against invading Gram-negative bacteria. To further confirm LRP-induced E. coli agglutination, bacterial cells were loaded with a fluorescent marker, Mitotracker Red CMXRos, and then incubated with rLRP. As expected, E. coli was specifically agglutinated in the presence of rLRP (Fig. 5B, a and c), but not in the absence of rLRP (Fig. 5 A, a and c).
|
|
To determine which region of LPS, i.e., O-chain, core oligosaccharide, or lipid A was recognized by LRP, inhibition of agglutination was examined using purified LPS from S and rough types of bacteria, LPS from core-deficient rough-type mutants of E. coli and S. enterica and free lipid A as inhibitors. As shown in Table I, rough-type LPS with the complete core of the E. coli core types R1, R2, R3, R4, and K-12 or with the Ra-core type of Salmonella were good inhibitors. S-type LPS exhibited varying results; some were good inhibitors (E. coli O4, O6, and S. enterica serovar abortus equi) and others showed low (E. coli O12) or no inhibitory ability (E. coli O15, O18, and S. enterica serovar Friedenau), indicating that LRP does not require the hypervariable O-polysaccharide chains for binding. The varying outcome is well understood because all S-type LPS preparations contain a fraction of varying size, where the core is not substituted with O-chains due to incomplete biosynthesis. LPS from core-deficient rough mutants of the chemotypes Rc and Rd2 were good inhibitors, Rb-type LPS was of low inhibitory activity, whereas Rd1- and Re-type LPS did not inhibit LRP-induced E. coli agglutination like free lipid A. These results suggest that the inner core oligosaccharide participates in the binding of LRP.
|
Furthermore, to re-examine the functions of LRP in naive plasma, we prepared the LRP-deficient plasma with the use of an affinity-purified LRP Ab and this plasma was confirmed by immunoblot analysis (Fig. 6A, lane 2) to be deficient of the 40-kDa LRP whereas naive plasma contained the LRP (Fig 6A, lane 1). When we incubated the LRP-deficient plasma with E. coli, agglutination activity was not induced (Fig. 6B), whereas agglutination was normal in naive plasma (Fig. 6C). This demonstrates that LRP is functioning as an LPS recognition protein to prevent the spreading of the invading bacteria in the insect body and is enhancing subsequent innate immune reactions, such as phagocytosis or encapsulation.
|
To ascertain the possible functions of LRP in early-stage defense reactions, E. coli were incubated in DFP-treated plasma or in the purified LRP to obtain bacteria coated with LRP. These bacteria were examined for the clearance rates compared with that of control bacteria not coated with LRP. The obtained results in Fig. 7 clearly demonstrate that the purified LRP-coated E. coli were more quickly cleared than those of E. coli coated with crude LRP. Also, we observed that the clearance of the nontreated E. coli is more rapidly cleared in the absence of the anti-LRP Ab than in the presence of the Ab. These results demonstrate that LRP is enhancing the clearance of bacteria. As a plausible function of LRP in naive larvae, we assumed that E. coli agglutination by LRP may enhance humoral or cellular innate immune processes. However, the fate of LRP-coated E. coli injected into the larvae was not determined in this study. There are three persuasive possibilities to explain how LRP-coated bacteria were killed in insect larvae. The first is that LRP-coated bacteria are easily phagocytosed and killed by putative professional phagocytes. Second, LRP-coated E. coli may induce the activation of the proPO system that produces the toxic quinone intermediates killing the invading microbes. Third, LRP-coated bacteria may stimulate the insect’s fat body or immune cells and then induce the secretion of antimicrobial peptides.
|
To determine the LPS-binding motif of LRP, we used CNBr to obtain the E. coli-agglutinating fragment from the intact LRP. LRP is connected by a 13-aa linker between the three EGF-like domains of the amino-terminal part and another three EGF-like domains at the carboxyl-terminal part (Fig. 3). Because LRP contains two Met residues in its amino acid sequence (Met174 and Met183), we anticipated that three or four fragments containing EGF-like domains would be generated from the purified LRP following CNBr treatment. As expected, we obtained four fragments when a CNBr-treated sample was analyzed in SDS-PAGE under reducing conditions (Fig. 8A, lane 2). These four bands were again separated on C18 reverse phase HPLC performance (Fig. 8B). After purification, we examined the E. coli-agglutinating activity of the purified peaks. Peak 3 containing bands 1 and 2, but not peaks 1 and 2, showed E. coli-agglutinating activity (Fig. 8C). When we determined the amino acid sequences of bands 1 and 2, both bands were determined as TTNPTAPRAY corresponding to the amino-terminal part (Thr1-Tyr10) of LRP. Under the same conditions, the sequences of bands 3 and 4 were determined as ICAPGFQ (Ile174-Gln180) and GSACEPL (Gly184-Leu190), respectively. These sequences are corresponding to the beginning of three EGF-like domains of the carboxyl-terminal part of LRP. These results strongly suggest that the three-EGF-like domains at the amino-terminal part, but not the carboxyl-terminal EGF-like domains, were involved in LPS-binding activity. As yet we do not know which specific EGF-like domain among the three EGF-like domains is involved in LPS recognition.
|
Because LRP are able to bind LPS, we investigated whether LRP could neutralize the LPS-induced mouse IL-6 production in mouse BMMC. It is well known that LPS induces the production of inflammatory cytokines (IL-1, TNF-
, IL-6, and IL-13) from mast cells (26). After a 30-min incubation of S-type or rough-type LPS (300 ng/ml) with LRP (150 ng/ml), LRP neutralized >90% and >75% of the LPS activity, respectively (Fig. 9).This clearly shows that LRP could interact with LPS at physiological conditions.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Recently, two studies identified two novel phagocytosis-mediating transmembrane receptors, Eater (33) and Draper (34) from Drosophila, which contain multiple EGF-like domain repeats. Interestingly, flies lacking the eater gene displayed normal responses in NF-
B-like Toll and immune deficiency signal pathways, but showed impaired phagocytosis and decreased survival after bacteria infection, suggesting that Eater is a major phagocytosis receptor for a broad range of bacterial pathogens in Drosophila (33). Also, the level of phagocytosis of apoptotic cells by Drosophila hemocytes was reduced when Draper expression was inhibited by RNA interference, indicating that Drosophila hemocytes execute Draper-mediated phagocytosis to eliminate apoptotic cells (34). Even though these two proteins are membrane-associated receptors, it would be interesting to examine whether the extracellular EGF-like region of Eater or Draper would bind to LPS or to determine which part of LPS can be recognized by these proteins.
To date, a number of LPS-recognition/binding molecules have been isolated and characterized from several animals, and these pattern recognition molecules have been shown to be involved in the innate immune reactions of both invertebrates and vertebrates. From vertebrates, enormous progress has been obtained in the elucidation of LPS recognition and signaling during mammalian phagocytes (35). According to a current model, recognition of LPS is initialized by the cooperative interplay between the LPS-binding protein, the membrane-bound and soluble forms of CD14, and the TLR-4-MD2 complex (36). In addition, the recognition of LPS leads to the activation of an intracellular signaling pathway, which results in the release of proinflammatory mediators (37). From invertebrates, LPS-binding proteins from cockroach Periplaneta americana (38), Factor C from horseshoe crab Tachypleus tridentatus (39), and LPS-binding lectin from Bombyx mori (25) were well characterized as a superfamily of LPS-recognition or -binding proteins and functioned in pathogen recognition and innate immunity in invertebrates. However, there is no report that a protein containing EGF-like domains functions as a soluble pattern recognition protein against LPS.
So far, several EGF-like domains containing seven-span transmembrane (EGF-TM7) receptors have been identified and expressed on human myeloid cells, including monocytes, macrophages, dendritic cells, and granulocytes (40). EGF-TM7 receptors belong to a subgroup of the long N-terminal family B G protein-coupled receptor-related TM7 receptor and consist of tandem repeats of amino-terminal EGF-like domains and a stalk region in the extracellular region. It was suggested (41) that the EGF-like domains of the EGF-TM7 receptors mediate cell-cell interaction by binding to specific cellular proteins and play an important role within the immune system. At present, however, the evidence for a specific immune function for EGF-TM7 receptors remains circumstantial. For example, although the ligand of some EGF-TM7 receptors, such as EMR2 and CD97, was identified as chondroitin sulfate, the ligand of most EGF-TM7 receptors has not yet been identified. Based on our studies where an EGF-containing LRP binds LPS, it may be possible that EGF-TM7 receptors may recognize endogenous carbohydrates containing -1,4-glucan linkages or exogenous LPS-like molecules. However, more careful studies will be needed before we can make any hypothesis regarding EGF-TM7 receptors.
In conclusion, we have shown that LRP containing EGF-like domains recognizes LPS and participates in the clearance of invaded bacteria.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by a research grant from the Korean Research Foundation (KRF-2004-041-C00278) and was also partially supported by National Research Laboratory Grant M10400000028-04J0000-02 (to B.L.L.) and a grant from Swedish Research Council (to K.S.). 
2 Address correspondence and reprint requests to Dr. Lee Bok Luel, National Research Laboratory of Defense Proteins, College of Pharmacy, Pusan National University, Geumjeong-gu, Busan, 609-735, Korea. E-mail address: brlee{at}pusan.ac.kr
3 Abbreviations used in this paper: PAMP, pathogen-associated molecular pattern; PGN, peptidoglycan; PGRP, PGN recognition protein; GRP, 1,3--D-glucan recognition protein; proPO, prophenoloxidase; LRP, LPS recognition protein; DFP, diisopropyl fluorophosphate; S, smooth; RT, room temperature; TEV, tobacco etch virus; EGF, epidermal growth factor; TM7, seven-span transmembrane; BMMC, bone marrow mast cell; PO, phenoloxidase.
4 The nucleotide sequences data reported in this article will appear in the DNA Data Bank in Japan, European Molecular Biology Laboratory, and GenBank nucleotide sequence database with the accession no. AB201466.
Received for publication January 26, 2006. Accepted for publication May 3, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-1,3-glucan: the binding domain and the cDNA cloning of -1,3-glucan recognition protein from the silkworm Bombyx mori. J. Biol. Chem. 275: 4995-5002. -1,3-glucan-binding protein from hemocytes of the freshwater crayfish Pacifastacus leniusculus: purification, characterization, and cDNA cloning. J. Biol. Chem. 275: 1337-1343. -D-glucan pattern recognition protein of Tenebrio molitor larvae that is specifically degraded by serine protease during prophenoloxidase activation. J. Biol. Chem. 278: 42072-42079. -D-glucan-dependent prophenoloxidase activation system of insect. J. Biol. Chem. 279: 3218-3227. -1,3-glucan pattern recognition and cellular encapsulation reactions. J. Immunol. 172: 2177-2185. This article has been cited by other articles:
|
|
 |
|
  K. Somogyi, B. Sipos, Z. Penzes, E. Kurucz, J. Zsamboki, D. Hultmark, and I. Ando Evolution of Genes and Repeats in the Nimrod Superfamily Mol. Biol. Evol., November 1, 2008; 25(11): 2337 - 2347. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and the purified LRP (
) mixed with the affinity-purified Ab against LRP (50 µg) were also injected into larvae. Five hundred microliters of hemolymph was collected after 30 min, and 1, 3, and 6 h from each larva, and then the number of viable bacteria in 100 µl of hemolymph was determined by plating. The experiment was repeated three times individually.
