Recombinant proteins and peptides for endotoxin biosensors, endotoxin removal, and anti-microbial & anti-endotoxin therapeutics

Abstract

Recombinant fragments of Factor C are disclosed. These proteins and peptides show great potency in recognizing, binding to, and neutralizing endotoxin. These molecules can thus be used for anti-microbial, anti-endotoxin, and anti-sepsis therapy. SSCrFCES is a 38 kDa protein representing the LPS-binding domain of Factor C. The ability of SSCrFCES to bind lipid A was analyzed using an ELISA-based assay as well as surface plasmon resonance. Surface plasmon resonance similarly carried out for SSCrFC-sushi-1,2,3-GFP, SSCrFC-sushi-1GFP, and SSCrFC-sushi-3GFP confirmed their superior affinity for endotoxin. The 50% endotoxin-neutralizing concentration of SSCrFCES against 200 EU of endotoxin is 0.069 μM, suggesting that SSCrFCES is an effective inhibitor of LAL coagulation cascade. Although partially attenuated by human serum, as low as 1 μM of SSCrFCES inhibits the LPS-induced secretion of hTNF-α and hIL-8 by THP-1 and human pheripheral blood mononuclear cells with a potency more superior than polymyxin B. SSCrFCES is non-cytotoxic, with a clearance rate of 4.7 ml/minute. The LD90 of SSCrFCES for LPS lethality in mice is achieved at 2 μM. These results demonstrate the endotoxin-neutralizing capability of SSCrFCES in vitro and in vivo, as well as its potential for use in the treatment of endotoxin-induced septic shock. Also embodied in this application is the use of the sushi peptides and their mutant derivatives as potent antimicrobials.

Description

FIELD OF THE INVENTION

[0002] The present invention relates to the use of recombinant polypeptides and synthetic peptides derived from a horseshoe crab Factor C as well as computationally designed peptide analogues, all of which have endotoxin-binding domain(s). The recombinant proteins may be expressed from insect cell clones, either as is or as fusion proteins, e.g. with green fluorescent protein (GFP). The extreme sensitivity of the present recombinant Factor C to LPS, with its unique LPS-binding domains which have unsurpassed binding affinity for LPS, may be exploited in accordance with the present invention for anti-endotoxin and anti-microbial therapeutics as well as for the tracing, detection, and removal of LPS or gram-negative bacteria. The present invention also relates to a method for treating bacterial infection of a subject by inducing bacteriostasis by administration of a recombinant Factor C protein.

BACKGROUND OF THE INVENTION

[0003] Endotoxin, also known as lipopolysaccharide (LPS), is an integral component of the gram-negative bacterial cell membrane and is responsible for many, if not all, of the toxic effects that occur during gram-negative bacterial sepsis (1). LPS is a mixture of glycolipids consisting of a variable polysaccharide domain covalently bound to a conserved glucosamine-based phospholipid known as lipid A. LPS directly stimulates host monocytes and macrophages to secrete a wide array of inflammatory cytokines that include tumor necrosis factor-o (TNF-α), interleukins-1 (IL-1), and interleukin-8 (IL-8) (2). Excessive release of these cytokines by host macrophages almost assuredly contributes to organ failure and death that occur after episodes of gram-negative bacterial sepsis (3). The proinflammatory bioactivities exhibited by LPS typically reside in the lipid A moiety (4).

[0004] LPS from gram-negative bacteria induces the amoebocytes of horseshoe crabs to aggregate and degranulate. Presumably, the LPS-induced coagulation cascade represents an important defense mechanism used by horseshoe crabs against invasion by gram-negative bacteria (5). The amoebocyte lysate constituted as the Limulus amoebocyte lysate (LAL) test has been used for decades as a tool for detecting trace concentrations of LPS in solution (6,7). The molecular mechanism of coagulation in horseshoe crab has been established and it involves a protease cascade. This cascade is based on 3 kinds of serine protease zymogens, Factor C, Factor B, proclotting enzyme, and one clottable protein, coagulogen (8). Being the initial activator of the clotting cascade, Factor C functions as a biosensor that responds to LPS.

[0005] Despite advances in antimicrobial therapy, septic shock and other clinical complications due to Gram-negative bacterial infections continue to pose a major problem. Endotoxin or lipopolysaccharide (LPS) present on the cell wall of Gram-negative bacteria (GNB) plays an important role in the pathophysiology of these infections. It does so by mediating toxicity and also mediating release of factors like tumor necrosis factor and interleukins (40), and also by forming a rigid shield around the bacteria protecting them from the effects of antibiotics. Therefore, the detection and/or removal of LPS from the bloodstream or any parenteral solution may aid in the prevention of the inflammatory and pyrogenic effects of LPS. The lipid A component of LPS plays the most important biological role; lipid A gives rise to all the ill effects elicited by endotoxin.

[0006] A number of LPS-binding proteins have been identified. Among them are the LPS binding protein, LBP (41), and bactericidal permeability increasing protein, BPI (18,42). LBP, a 60 kDa mammalian serum protein, has a binding site with a high degree of specificity for lipid A (43). BPI, a 55 kDa protein found in human neutrophils, is capable of binding to the toxic lipid A moiety of LPS resulting in neutralization of the endotoxin (18,42,44,45).

[0007] The circulating amoebocytes of the horseshoe crab contain an array of proteins that are capable of binding and neutralizing LPS. The Limulus antilipopolysaccharide factor, LALF, an 11.8 kDa LPS-binding peptide, has been identified in the amebocytes of horseshoe crabs Limulus polyhemus and Tachypleus tridentatus. LALF has subsequently been isolated and characterized (46-49). Purified LALF has been shown to bind LPS and exhibit endotoxin neutralization (50,19,51,52). Two other LPS-binding proteins from horseshoe crab hemocytes are tachyplesin (53,54) and big defensin (55).

[0008] Factor C is a serine protease zymogen. It is the key enzyme in the C. rotundicauda amoebocyte lysate (CAL) that is activated by LPS to initiate the coagulation cascade (56-58). Factor C activity is the basis of a very sensitive assay for femtogram levels of endotoxin used in the quality control of pharmaceutical products (59). The importance of Factor C in the detection of endotoxin has thus led to the expression of recombinant Factor C, rFC (12,60,61,73-38), as an alternative source that should alleviate the batch-to-batch and seasonal variation in the sensitivity of detection of endotoxin which is a recognized drawback with conventional amoebocyte lysate (59-61).

SUMMARY OF THE INVENTION

[0009] Since Factor C can be activated by femtograms of LPS, it is thought that Factor C has an LPS-binding region that exhibits exceptionally high affinity for LPS. Consequently, this LPS-binding domain can be utilized to detect and remove pyrogenic contaminants in pharmaceutical products intended for parenteral administration as well as for in vivo immunohistochemical determination of endotoxin localization (9).

[0010] The LPS-binding property of Factor C resides in the amino-terminal region spanning 333 amino adds. This short region constitutes a signal peptide, a cysteine-rich region, followed by epidermal growth factor-like domain and finally 3 sushi domains. High LPS affinity, comparable to the native Factor C, requires the correct formation of 9 disulfide bonds (16). This obstacle is compounded by the presence of a cysteine-rich region. Here, for the first time, we report the expression and secretion of a functional LPS-binding domain of C. rotundicauda Factor C (SSCrFCES) via a novel secretory signal. The secretory signal is disclosed in U.S. patent application Ser. No. 09/426,776, filed Oct. 26, 1999. The entire disclosures of 09/426,776 and of the provisional application upon which it is based, 60/106,426, are hereby expressly incorporated by reference.

[0011] Homologous Factor C zymogen cDNAs have been cloned from two of the four extant species of horseshoe crab, Carcinoscorpius rotundicauda (CrFC) (10). Initial attempts to express CrFC and its truncated forms in E. coli resulted in a non-active enzyme (11). Subsequently, CrFC was cloned and expressed in Saccharomyces cerevisiae and a methylotropic yeast, Pichia pastoris. However, neither the Factor C nor the Saccharomyces cerevisiae a mating factor signal sequences were capable of directing secretion of the recombinant protein into the culture media for easier purification (12). Full-length CrFC expressed in yeast was not enzymatically active although it retained endotoxin-binding properties (13).

[0012] Expression in a baculoviral system (U.S. patent application Ser. No. 09/081,767, filed May 21, 1998) yielded recombinant Factor C (rFC) with LPS-inducible enzyme activity. The entire disclosures of 09/081,767 and of the provisional application upon which it is based, 60/058,816, are hereby expressly incorporated by reference. The rFC has extremely high sensitivity to trace levels of LPS (<0.005 EU/ml). Before these experiments, the LPS-binding domain of Factor C exhibiting high affinity for LPS was never before successfully expressed in a heterologous host. The difficulty in doing so was largely due to its highly complex mosaic structure. While many highly disulfide-bonded proteins, like epidermal growth factor (14) and secreted acetylcholinesterase (15), were successfully expressed, few display the kind of complexity posed by the Factor C LPS-binding domain.

[0013] A form of SSCrFCES was secreted in accordance with the present invention and was purified to homogeneity. The biological functions of the recombinant SSCrFCES were assessed by measuring the ability of the SSCrFCES to bind lipid A using an ELISA-based lipid A binding assay as well as surface plasmon resonance interaction. Other subfragments containing the LPS-binding domain(s)—e.g., SSCrFCsushi-1,2,3-GFP, SSCrFCsushi-1-GFP, SSCrFCsushi-3-GFP (fusion constructs with green fluorescent protein, GFP)—as well as synthetic peptides, e.g.; sushi-1 (S1), sushi-1Δ (S1Δ), sushi-3 (S3), and sushi-3Δ (S3Δ), each of 34 mer length, and designed variant forms of peptides bearing BHBHB and/or BHPHB (where B=basic, H=hydrophobic, P=polar amino acids)—also show strong affinity for endotoxin.

[0014] The ability of these proteins and peptides to mediate inhibition of endotoxin-induced Limulus amoebocyte lysate (LAL) coagulation was measured with a sensitive LAL Kinetic-QCL assay. The SSCrFCES protein and peptides were also tested for their ability to suppress LPS-induced cytokines TNF-α and IL-8) produced by THP-1 and normal human peripheral blood mononuclear cells (hPBMC). SSCrFCES and the peptides were non-cytotoxic. SSCrFCES has a clearance rate of 4.7 ml/min. We also show that low doses of SSCrFCES protein and the synthetic peptides protect galactosamine-sensitized mice from LPS-induced lethality. The peptides have strong antimicrobial potencies and can therefore be used as potent therapeutics.

[0015] The present invention thus includes treating bacterial infections by administration of proteins that will bind to endotoxins, especially endotoxins produced by gram-negative bacteria, to an infected subject. The binding is apparently mediated by the lipid A component of the endotoxin. The administered protein induces bacteriostasis (that is, inhibition of bacterial proliferation) in the subject.

[0016] A preferred embodiment of this aspect of the invention is one wherein recombinant Factor C is the administered protein. The recombinant Factor C can be a full-length Factor C protein, or any portion thereof that retains the activity of binding to lipid A. It is not necessary that the Factor C retain its serine protease enzymatic activity for the protein to be effective in the method of the invention. It may in fact be beneficial if the serine protease activity is absent.

BRIEF DESCRIPTION OF THE DRAWINGS

[0017] FIG. 1(A). Coomassie brillant blue-stained 12% reducing SDS-PAGE profile of crude and purified SSCrFCES. The recombinant protein, SSCrFCES, was effectively secreted into the culture medium of S2 cells and identified as a 38 kDa protein band. Purification using ISOPrime™ resulted in an isoelectrically homogenous SSCrFCES.

[0018] FIG. 1(B). Immunoblotting analysis was perfomed with INDIA™ His-HRP antibody and visualized using SuperSignal™ Chemiluminescence. A specific 38 kDa band, in close agreement to calculated SSCrFCES size, was identified as the only secreted and purified protein harbouring a poly-histidine tag. Exposure time, using Biomax™ film (Kodak), was limited to 5 sec. Lanes are identified as follows: 1, Low-Molecular Weight marker (Pharmacia); 2, control medium (30 μg); 3, crude SSCrFCES medium (30 μg); 4, Affinity purified SSCrFCES (1 μg); 5, ISOprime™ purified SSCrFCES (1 μg)

[0019] FIG. 2(A). SSCrFCES displayed a biphasic binding profile to lipid A measured by an ELISA-based assay. Three different concentrations of lipid A were coated overnight onto Polysorp™ plates (Nunc). Varying concentrations of SSCrFCES were allowed to interact with the immobilized lipid A. The amount of bound SSCrFCES was determined by rabbit anti-SSCrFCES IgG and quantitated by ABTS substrate. The O.D.405 nm of the samples and reference wavelength at 490 nm were determined using a microtiter plate reader. The biphasic response is indicative of multiple binding sites for lipid A.

[0020] FIG. 2(B). SSCrFCES binds to lipid A at a stoichiometry of ˜3 lipid A molecules per SSCrFCES. A plot of the molar ratio of bound SSCrFCES to immobilized lipid A, gave a value of 0.37 at saturation. This means that each SSCrFCES molecule has the ability to bind ˜3 lipid A molecules.

[0021] FIG. 2(C). A Hill's plot showing Hill's coefficient, determined by the slope of the straight line obtained from plotting that data according to the Hill's equation, is 2.2. This indicates that SSCrFCES exhibited positive cooperativity in lipid A binding.

[0022] FIG. 3(A). A surface plasmon resonance (SPR) sensogram depicting the interaction of SSCrFCES, with immobilized lipid A. 800 ng/100 μl of SSCrFCES was injected which resulted in an increase of 200 relative response unit. After the dissociation phase, by passing PBS in a running buffer, INDIA™ His-HRP antibody was injected. The further increase in relative response unit clearly indicates that SSCrFCES is bound to lipid A. The surface was regenerated by a pulse of 100 mM NaOH. At all times, the flow rate was maintained at 10 μl/min.

[0023] FIG. 3(B) is a sensogram depicting the interaction of SSCrFCsushi-1,2,3-GFP with immobilized lipid A.

[0024] FIG. 3(C) is a sensogram depicting the interaction of SSCrFCsushi-1-GFP with immobilized lipid A.

[0025] FIG. 3(D) is a sensogram depicting the interaction of SSCrFCsushi-3-GFP with immobilized lipid A.

[0026] FIG. 3(E) is a sensogram depicting the interaction of certain synthetic peptides with immobilized lipid A. The table (inset to E) shows the binding properties of the synthetic peptides to lipid A.

[0027] FIG. 4(A). SSCrFCES increases LAL-based Kinetic QCL reaction time. Various concentrations of SSCrFCES were incubated with 200 EU/ml of E. Coli (055:B5) LPS for 1 h at 37° C. Following pre-incubation, the mixture was diluted 1000-fold prior to assay by Limulus Kinetic-QCL. The O.D.405 nm of each well of the microtitre plate was monitored at time intervals of 5 min over a period of 2 h. The endotoxin-neutralizing concentration (ENC50) of SSCrFCES, which is the concentration of SSCrFCES that increase the mean reaction time by 50% was found to be 0.069 μM. Mean reaction time using only LPS is designated as 0%.

[0028] FIG. 4(B). Binding of S1, S1Δ, S3, and S3Δ to LPS. The 50% endotoxin-neutralising concentration (ENC50) were determined to be S1=2.25 μM, S1Δ=0.875 μM, S3=1 μM, and S3Δ=0.625 μM.

[0029] FIG. 4(C). Hill's plot for interaction between synthetic peptides and lipid A shows that S1 exhibits high positive co-operativity of n=2.42, indicating that more than 2 S1 peptides interact with 1 LPS molecule.

[0030] FIG. 5(A). SSCrFCES inhibits LPS-induced hTNF-α secretion from THP-1 in a dose-dependent manner. PMA-treated THP-1 cells were treated with 25 ng/ml of E. coli 055:B5 LPS which were preincubated with varying concentrations of SSCrFCES. After 6 h of stimulation, the culture medium was assayed for TNF-α. The decrease in TNF-α were expressed as percentage of control (LPS only). Complete inhibition of TNF-α was achieved using 1 μM of SSCrFCES.

[0031] FIG. 5(B). SSCrFCES inhibits LPS-induced hIL-8 secretion from THP-1 in a dose-dependent manner. PMA-treated THP-1 cells were treated with 100 ng/ml of E. coli 055:B5 LPS which were preincubated with varying concentrations of SSCrFCES. After 6 h of stimulation, the culture medium was assayed for IL-8. The decrease in IL-8 secretion was expressed as percentage of control (LPS only). 95% inhibition of IL-8 secretions were achieved using 1 μM of SSCrFCES.

[0032] FIG. 6(A). The ability of SSCrFCES to inhibit LPS-stimulated TNF-α secretion from PBMC cells. In the absence of human serum, addition of only 8.5 nM of SSCrFCES caused 50% inhibition of TNF-α response to 10 ng/ml LPS. SSCrFCES pre-incubated with 10% human serum required 17-fold more protein to achieve 50% inhibition. The attenuation can be minimized if the SSCrFCES was mixed with endotoxin 5 min before the addition of serum, thus requiring only 4-fold more SSCrFCES for 50% inhibition of cytokine release.

[0033] FIG. 6(B). The ability of SSCrFCES to inhibit LPS-stimulated IL-8 secretion from PBMC cells. In the absence of human serum, addition of only 8.5 nM of SSCrFCES caused 50% inhibition of IL-8 response to 10 ng/ml LPS. SSCrFCES pre-incubated with 10% human serum required 17-fold more protein to achieve 50% inhibition. The attenuation can be minimized if the SSCrFCES was mixed with endotoxin 5 min before the addition of serum, thus requiring only 4-fold more SSCrFCES for 50% inhibition of cytoidne release.

[0034] FIG. 6(C). The ability of synthetic peptides to suppress LPS-induced TNF-α.

[0035] FIG. 7. SSCrFCES is not cytotoxic to mammalian cells. At the highest concentration of 4 mg/ml or 109 μM, only 20% cell lysis was observed.

[0036] FIG. 8. Pharmacokinetic analysis of SSCrFCES shows that clearance rate of biotin-labeled SSCrFCES in C57BL/6J mice is 4.7 ml/min.

[0037] FIG. 9(A). SSCrFCES protects C57BL/6J mice against LPS-induced lethality. 100% LPS-induced lethality was achieved using 2.0 ng of E. coli 055:B5 within 7 h. The percentage of survival was increased to >90% when 2 and 4 μM of SSCrFCES were injected i.v. 10 min after LPS challenge. Kaplan-Meier analysis indicates that there is significant difference between 1 μM and 2 μM of SSCrFCES (P<0.0005). No significant difference was observed between 2 μM and 4 μM of SSCrFCES.

[0038] FIG. 9(B). S1, S1Δ, S3, S3Δ, and other designed variant peptides protect C57BL/6J mice against LPS-induced lethality. 100% LPS-induced lethality was achieved using 2.0 ng of E. coli 055:BS within 7 h. The synthetic peptides (25 or 75 μg) were pre-incubated with LPS for 30 min prior to i.p. injection. S1, S1Δ, and S3 conferred 20-55% decrease in LPS-induced lethality. However, S3Δ is significantly more effective in protection, where 75 ug was sufficient to confer 100% protection.

[0039] FIG. 10(A). CrFC21 (SEQ ID NO:4) showing functional domains of Factor C.

[0040] FIG. 10(B). Recombinant fragments: ssCrFCES; sushi-1,2,3-EGFP; sushi 1-EGFP; and sushi-3-EGFP fusion proteins. Sushi peptides of 34 mer each (S1, S1Δ, S3, & S3Δ).

[0041] FIG. 11A. Sequences of V1 and V2 peptides.

[0042] FIG. 11B. Sequences of peptides featured in Table 3.

[0043] FIG. 12. The microbicidal concentrations (MBC) of sushi peptides against P. aeruginosa (ATCC 27853). An initial density of 105 cfu/ml of P. aeruginosa was used in the assay. Doubling of the peptide concentrations in the region of 0.03-0.5 ug/ml resulted in exponential reduction of bacterial count. S1 and S3 were more effective against P. aeruginosa than S1Δ and S3Δ.

[0044] FIG. 13. Time-dependent killing of P. aeruginosa ATCC 27853. An initial density of 109 cfu/ml of P. aeruginosa was used in the assay. The effect of test peptides at 0.06 ug/ml was assessed by enumerating the viable (cfu/ml) at indicated time intervals after overnight incubation. The bacterial count was exponentially reduced to achieve MBC90 within 7 min. By 30-40 min, the bacterial was completely eradicated.

[0045] FIG. 14. Drop count plates for the killing rate assay at 0.06 ug/ml of sushi peptides in general, monitored at the indicated time intervals. Segments of the plates contain P. aeruginosa culture at 10-fold dilution starting from 10−1 to 10−8 from the upper quadrant in anti-clockwise direction. S3Δ peptide eradicated the bacteria at a log reduction rate.

[0046] FIG. 15. Electron micrographs showing examples of how the antimicrobial peptides kill the bacteria.

[0047] FIG. 16. Sushi peptides display negligible hemolytic activities. Human and rabbit erythrocytes at 0.4% were reacted separately with different doses of peptides (6-100 μg/ml). 0.4% erythrocytes lysed in 1% Triton-X was taken as 100% lysis. The negative control was 0.4% erythrocytes in pyrogen-free saline. Sushi peptides were minimally hemolytic up to concentration of 100 μg/ml. S1, S1Δ, and S3 showed negligible haemolysis and S3Δ caused a 35% haemolysis at 100 μg/ml. Concentration of peptide to induce 50% haemolysis: S1 290 μg/ml; S1Δ 295 μg/ml; S3 160 μg/ml; and S3Δ 120 μg/ml.

[0048] FIG. 17. Example of S3Δ-peptide coupled Agarose CL-6B beads bound with FITC-LPS, seen under microscope. (A) Bright field observation; (B) Beads with FITC-LPS bound, seen under UV light; (C) Bound beads after treatment with 1% DOC—no FITC-LPS left on the beads (observed under UV light).

[0049] FIG. 18. A test of binding conditions of LPS to S3Δ peptide affinity beads under increasing pH and ionic strength. (A) pH of 4.0 and 5.0 (in 20 mM sodium acetate), pH 6.8 and 9.1 (20 mM Tris-HCl). All buffers were supplemented with 50 mM NaCl. (B) Different ionic strength: 20 mM Tris-HCl (pH 6.8) were supplemented with different concentrations of NaCl, except of the 0 mM point which contained pyrogen-free water as control.

[0050] FIG. 19. Immunoblot showing expression of rFC (pHILD2/CrFC21; lane 1), rFCEE (pHILD2/CrFC21EE; lane 2) and rFCSN (pPIC9/CrFC26SN; lane3) in the crude supernatant. Arrows indicate the immunoreactive recombinant Factor C proteins: 132 kDa full-length rFC, 90 kDa truncated rFCEE and 89 kDa truncated rFCSN. The molecular weight markers (MW) are labeled in kDa.

[0051] FIG. 20. SDS-PAGE analysis showing the protein profiles of the different preparations of rFC: crude supernatant (lane 1); (NH4)2SO4 precipitated sample (lane 2); Biomax™-50 enriched rFC (lane 3); and Sephadex™ G-100 purified sample (lane 4). Ten micrograms of each protein sample were loaded. Arrow indicates the 132 kDa full-length rFC. The molecular weight markers (MW) are labeled in kDa.

[0052] FIGS. 21A and 21B. Modified Western blot to show binding of Factor C to LPS strips (FIG. 21A) and lipid A strips (FIG. 21B). Lanes 1: crude rFC; 2: (NH4)2SO4 precipitated rFC; 3: Biomax™-50 purified rFC; 4: Sephadex™ G-100 purified rFC; 5: Biomax™-50 purified rFCEE; 6: Biomax™-50 purified rFCSN; 7: pHILD2/151 supernatant. The 7-20 kDa lipid A bands are indicated between the 2 arrows.

[0053] FIG. 22A. Competitive effects of 50, 100 and 200 μg total protein of crude rFC on LPS-mediated activity of CAL Factor C enzyme activity. Dashed line illustrates the ratio of crude rFC to LPS (1000:1) for a percentage competition of >80%. Results are the means±S.D. of three independent experiments.

[0054] FIG. 22B. Competitive effects of 50, 100 and 200 μg Biomax™-50 enriched rFC on LPS-mediated activity of CAL Factor C enzyme activity. Dashed line illustrates the ratio of rFC to LPS (100:1) for a percentage competition of >80%. Results are the means±S.D. of three independent experiments.

[0055] FIG. 23. Comparison of the competitive efficiencies of full-length rFC and truncated rFCEE on LPS-mediated enzymatic activity of CAL Factor C. Each protein sample (rFC or rFCEE) was enriched by Biomax™-50 ultrafiltration, and 100 μg was used in the competition assay. The percentage competition was obtained after normalization with the background competition by rFCSN. Results are the means±S.D. of three independent experiments.

[0056] FIG. 24. Interactive binding of rFC to immobilized lipid A in a BIACORE X™ sensor. Lipid A (100 μg/ml) was immobilized on the sensor chip. The respective protein samples were flowed through and relative responses recorded in response units (RU) by the BIACORE X™ instrument. Plateaus 1A, 2A and 3A on the sensorgram represent the relative responses of Biomax™-50 enriched rFCSN, rFCEE and rFC, respectively, to immobilized lipid A. Arrows show the RU due to regeneration with 0.1M NaOH. Inset shows the net percentage RU of rFC and rFCEE to immobilized lipid A. The percentage RU of each protein sample was calculated based on the relative RU of the protein sample and that of immobilized lipid A. The net RUs of rFC and rFCEE were obtained after normalizing their relative RUs with that of rFCSN.

[0057] FIG. 25. The bacteriostatic effects of Sephadex™ G-100 purified rFC on the growth of the Gram-negative bacteria: E. coli, K. pneumoniae, P. aeruginosa, and S. typhimurium. rFC was most efficacious against K. pneumoniae whereas the bacteriostatic activity against P. aeruginosa declined rapidly after 4 h.

[0058] FIGS. 26A-26E. Agglutination of E. coli by rFC (FIG. 8A) and rFCEE (FIG. 8B). Observations were made with a Nikon MICROPHOT™-FXA microscope (400× magnification). No agglutination was seen with rFCSN (FIG. 8C), pHILD2/151 (FIG. 8D) and 0.85% saline (FIG. 8E).

[0059] FIG. 27. The protective effect of 10 μg rFC purified through Sephadex™ G-100 on actinomycin D sensitized/LPS-challenged mice. Pre-incubation of LPS with rFCSN did not confer protection of mice against the endotoxic effects of LPS. For comparison, rFC conferred 60-70% protection.

[0060] FIG. 28. Line drawings of rFC (full length) and its deletion homologues, given with their corresponding start and end amino acid positions based on the CrFC21 clone (SEQ. ID. NOs. 3 and 4, U.S. Pat. No.: 5,716,834). Amino acid residues are numbered as in SEQ. ID. NO. 4. rFC, rFCEE, rFCES(sushi-1,2,3), rFC(sushi-1) and rFC(sushi-3) have endotoxin-binding site(s). Sushi () domains 1, 2, and 3 denote secondary structures in Factor C, with ‘sushi-like’ folding patterns. rFCSN does not contain any endotoxin-binding site. The lines are not drawn to scale.

[0061] FIG. 29. Binding of rFC produced in baculovirus-infected Sf9 cells to LPS from various bacteria.

[0062] FIG. 30. Bacteriostasis induced by rFC produced in baculovirus-infected Sf9 cells in cultures of different Gram-negative bacteria.

[0063] FIG. 31. Protection of mice from LPS lethality by administration of rFC produced in baculovirus-infected Sf9 cells.

BRIEF DESCRIPTION OF TABLES

[0064] Table 1 presents a comparison between binding affinity for lipid A of Factor C-derived sushi proteins and other LPS-binding proteins.

[0065] Table 2 presents a comparison of MBC50, MBC90, hemolytic activity, and cytotoxic activity of sushi and other cationic peptides on test microorganisms.

[0066] Table 3 provides indicators of LPS-binding, anti-LPS, and antimicrobial activities of Factor C and various peptides. In Table 3, column I shows affinity for LPS binding of peptide to Lipid A immobilized on an HPA chip, column II shows Hill's Coefficient—the stoichiometry of binding of the number of peptide molecules to 1 LPS molecule, column III shows Circular Dichroism (CD) analysis of peptide structures in the presence of 0.75 nM lipid A (αH: α-helical; β: β-sheet; T: turn; R: random), column IV shows neutralization (EC50)—μM of peptide needed to neutralize 50% of 200 BU/ml of LPS-induced LAL reaction, column V shows the amount of peptide needed to cause 50% suppresion of LPS-induced cytokine release (TNF-α), column VI shows mouse protection assays—2 ng LPS pre-incubated with peptide for 30 minutes before injection into C57/BL, column VII shows cytotoxicity (cell lysis) assays—for S4-S9: EC50=[peptide] to cause 50% lysis cytotoxicity, column VIII shows hemolytic activity at 100 μg peptide, and column IX shows MBC90 (microbiocidal concentration of peptide that kills 90% of bacteria) or MIC90 (minimal inhibitory concentration of peptide that inhibits 90% of bacteria).

DETAILED DESCRIPTION OF THE INVENTION

[0067] The present invention provides efficient, high affinity recombinant proteins and peptides for gram-negative bacterial endotoxin. These molecules can be used, among other things, for: (a) anti-microbial, anti-endotoxin, anti-sepsis therapeutics; (b) tracing and detection or localization of gram-negative bacteria via, for example, the GFP component of SSCrFCsushi-GFP fusion proteins; and (c) development of LPS-specific affinity chromatography systems to purify endotoxin-contaminated samples or biological fluids.

[0068] The present invention lies in part in methods for treating bacteremia using proteins that bind to bacterial endotoxin as a therapeutic agent. A particularly effective protein is a recombinant Factor C protein, or any portion thereof that retains the biological activity of binding to lipid A.

[0069] cDNAs encoding Factor C proteins from Carcinoscorpius rotundicauda have been previously described (10,73). Recombinant Factor C from Carcinoscorpius rotundicauda (rCrFC) has been produced in vitro by coupled transcription/translation systems. However, the present invention resides partly in the development of in vivo systems, especially using insect cells as the host cell, for efficient production of rFC by expression of cloned DNA.

[0070] Also, the protection of rFC from activation and subsequent self-proteolysis by binding of endotoxin which may be present in solutions used in isolation of the protein is described in U.S. Pat. No. 5,716,834, the entire disclosure of which is hereby incorporated by reference. Basically, dimethylsulfoxide (Me2SO or DMSO) is added to solutions which are used during the purification process. Even greater protection of the rFC is achieved by also adding an agent effective for chelating divalent metal ions to the purification solutions.

[0071] cDNAs appropriate for expression in the presently-described system can be cDNAs encoding Factor C of any horseshoe crab. Two representative nucleotide sequences are presented as SEQ ID NO:1 and SEQ ID NO:3 (encoding the amino acid sequences of SEQ ID NOs:2 and 4). A composite DNA sequence, assembled from incomplete cDNA fragments, encoding the Factor C of Tachypleus tridentatus is disclosed by Muta et al (49).

[0072] Factor C appropriate for use in the present invention can be produced by any method typical in the art, but is preferably made in a eukaryotic host cell. Production of rFC in yeast host-vector systems is described in reference 75. Recombinant Factor C produced in yeast is found to lack serine protease activity, but, as shown in the working examples below, protein produced in yeast is still effective in both lipid A and endotoxin binding and in inducing bacteriostasis. Production of rFC in yeast host-vector systems is described in detail in co-pending U.S. patent application Ser. No. 08/877,620. Recombinant Factor C for use in the invention can also be produced by a baculovirus host-vector system or in another suitable insect cell host-vector system, such as one for Drosophila cells. Copending U.S. patent applications Nos. 09/081,767, 60/106,426 and 09/201,786 provide detailed description of production of rFC in such systems.

[0073] The endotoxin/lipid A-binding domain of Factor C lies within the amino terminal portion of the protein encompassed by rFCES; that is, the first 350 amino acids, numbered as in SEQ. ID. NO. 4. Referring to FIG. 28, endotoxin/lipid A binding activity is found in the truncated rFCEE (amino acids 1-766), rFCES (amino acids 29-330), rFC(sushi-1) (amino adds 29-201) and rFC (sushi-3) (amino acids 264-330). Molecular modeling studies suggest that the contacts are made by portions of the protein lying in the cysteine-rich domain, especially amino acids 60-70, in the sushi-1 domain, especially amino acids 170-185 and in the sushi2 domain, especially amino acids 270-280. Thus, a protein having at least these three portions of Factor C, which can be joined by a random amino acid sequence or by other chemical linkage, is expected to be useful in the method of the present invention.

[0074] As noted above, naturally-occurring Factor C proteins, and rFC that is full-length and produced in baculovirus-infected or other insect cell lines, possesses a serine protease activity. That activity is activated by endotoxin or lipid A binding. It might be found that the serine protease activity of the rFC produces undesired side effects when treating a subject with rFC according to the invention. Thus, in preferred embodiments of the present invention, the serine protease activity of the rFC is inactivated, either chemically or by mutation, or the domain providing that activity is deleted from the protein.

[0075] The portion of Factor C from horseshoe crab that constitutes the serine protease domain is approximately from amino acid 760 to the carboxy terminus of the protein, numbered as in SEQ. ID. NO.:4. Furthermore, the particular amino acids that constitute the catalytic residues are His809, Asp865, and Ser966. Thus, inactivation of these residues by chemical modification or by site-specific mutation can be used to provide rFC that will bind to lipid A, but lacks serine protease activity.

[0076] Chemical modifications to inactivate serine protease activity are well-known in the art. Methods for introducing site-specific mutations into any particular polypeptide are also well-known in the art.

[0077] Colorimetric and fluorescent assays for the serine protease activity of rFC are described in detail in co-pending application Ser. No. 09/081,767, the entire disclosure of which is hereby incorporated by reference. These assays are appropriate for screening mutant forms of rFC for serine protease activity. Assays for lipid A and endotoxin binding is also described in co-pending application Ser. No. 09/081,767 that can be used to ascertain that the serine protease-deficient mutant retains the lipid A/endotoxin binding activity required if the protein is to be used in the present invention.

[0078] “Stringent conditions” for hybridization are those that provide for hybridization of sequences having less than 15% mismatch, preferably less than 10% mismatch, most preferably 0% to 5% mismatch. Exemplary of such conditions, using probes of 50 bases or longer, are an aqueous solution of 0.9 M NaCl at 65° C.; an aqueous solution of 0.98 M NaCl, 20% formamide at 42-45° C. The conditions will vary according to the length of the probe, its G+C content and other variables as known to the skilled practitioner (54). Exemplary wash conditions following hybridization are an aqueous solution of 0.9 M NaCl at 45-65° C., preferably 55-65° C. Lower salt, or addition of an organic solvent such as formamide, in the wash buffer will increase the stringency of the condition as known in the art.

[0079] A preferred hybridization condition is at 42° C. in 50% formamide, 5×SSC, 1× Denhardt's solution, 20 mM phosphate buffer, pH 6.5, 50 μg/ml calf thymus DNA, 0.1% SDS. Salt and temperature conditions equivalent to the hybridization conditions employed can be calculated from the following equation:

T m =81.5° C.−16.6(log10[Na+])+0.41(% G+C)−0.63(% formamide)−(600/l)

[0080] where l=the length of the hybrid in base pairs.

[0081] A preferred washing condition is in 1×SSC, 0.1% SDS washing solution at room temperature, followed by washing at high stringency with 0.1×SSC, 0.1% SDS at 42° C. and 2× with 0.1×SSC/0.1% SDS for 15 min. each at 42° C.

[0082] Preferred versions of rFC for use in the method of the invention are those encoded by polynucleotides that will hybridize to a nucleic acid having the sequence of SEQ. ID. NO. 1 or SEQ. ID. NO. 3 under stringent conditions. Most preferred versions of rFC are those having the amino acid sequence of SEQ. ID. NO. 2 or SEQ. ID. NO. 4.

[0083] For administration to a subject for treatment of bacterial infection or to induce bacteriostasis, the rFC is formulated with pharmaceutically acceptable carriers appropriate for the route of administration. Formulation of polypeptides for administration is known in the art; the practitioner is referred, for example, to reference 79. The route of administration is not particularly limiting of the invention, but preferred routes are intraperitoneal intravenous administration.

[0084] The proteins for administration are preferably formulated in pharmaceutical saline solutions such as 0.9% saline, phosphate buffered saline and the like. The polypeptides can be provided in lyophilized form and reconstituted for administration. The final concentration of the protein in the formulation administered is one that would provide a suitable dosage as described below.

[0085] Polypeptide therapeutic agents are known to be susceptible to degradation in the body, usually due to the action of proteolytic enzymes. Thus, the rFC administered according to the present invention might desirably be derivatized to inhibit such degradation. For example, carboxy-terminal amidation of the protein is known in the art to inhibit degradation by proteases present in serum. Particular derivations of proteins to improve their resistance to degradation in vivo and methods for accomplishing them are well-known in the art.

[0086] The dosage to be administered will of course be tailored to the particular form of rFC administered and the route of administration. Tailoring of dosage is considered within the skill of the routine practitioner. A dosage within the range 0.01 to 1.0 mg/kg body weight is acceptable; preferably the dosage will be within the range of 0.1 to 1.0 mg/kg, most preferably in the range of 0.3 to 0.4 mg/kg.

[0087] Doses may be administered either by bolus or by infusion. The particular rate of administration will be determined partly by the half-life of the protein in the body, which will be influenced by the particular structure of the protein and also by the route of administration. Assessment of pharmacokinetics necessary to determine the precise rate and dosage of the particular protein to be administered is considered within the skill of the practitioner.

[0088] The following exemplary embodiments of the invention serve to illustrate the invention. The examples are not to be considered limiting of the scope of the invention, which is defined only by the claims following.

EXAMPLE 1

Purification of Stably Expressed and Secreted Recombinant SSCrFCES

[0089] Stable cell lines of Drosophila S2 clones expressing SSCrFCES (U.S. patent application Ser. No. 09/426,776) were routinely cultured in serum-free DES Expression medium and maintained at 25° C. in a humidified incubator.

[0090] (a) Purification of SSCrFCES using a TALON Column

[0091] The medium containing SSCrFCES was initially concentrated and desalted via 3 rounds of ultrafiltration using a 10 kDa cutoff membrane in an Amicon stirred cell (Millipore). Affinity chromatography purification under denaturing conditions yielded a 38 kDa protein of interest, in addition to a 67 kDa protein. Western blot analysis indicated that the 67 kDa protein does not contain the carboxyl poly-His tag. Thus this larger protein is likely due to non-specific adsorption to the resin.

[0092] (b) Purification of SSCrFCES by Preparative Isoelectric Membrane Electrophoresis

[0093] Typically, 2 liters of conditioned medium were initially subjected to successive ultrafiltration using a 100 kDa and 10 kDa molecular weight cutoff with the Pellicon system (Millipore). The medium was concentrated seven-fold. The enriched SSCrFCES was purified to isoelectric homogeneity using Preparative Isoelectric Membrane Electrophoresis (Hoefer IsoPrime™, Pharmacia). The pI of the SSCrFCES was determined to be 7.1 at 4° C. A set of four membranes were made, with pHs of 6.5, 7.0, 7.25, and 7.5. The concentration of acrylamido buffers used for the membranes were calculated based on information in Righetti and Giaffreda (17). The four membranes were assembled in order, from acidic to basic, to delimit five chambers. Each sample reservoir vessel was filled with 30 ml of pyrogen-free water and pre-run at 4° C. at 4 Watts constant power (3000 V limiting, 20 mA maximum) for two hours.

[0094] After removing the pre-run water, the protein sample was placed in sample reservoir vessel corresponding to the chamber delimited by pH 7.0 and 7.25. The IsoPrime was conducted under the same conditions for 34 days without detrimental effect on the protein, and the content from each chamber was analyzed on a 12% SDS-PAGE. The scheme reported here has been found to be reproducible in our laboratory throughout the course of approximately two years. The overall recovery of SSCrFCES binding capacity is nearly 90%. This is attributable to its extreme stability conferred by the presence of 9 disulfide bonds.

[0095] (c) Analysis of the Purified SSCrFCES

[0096] The distribution of the protein was identified using Chemiluminescent Western blot. SDS-PAGE analysis of Drosophila cells transformed with the recombinant vector is shown in FIG. 1A. The Western blot revealed the presence of a protein with an apparent molecular weight of 38 kDa (FIG. 1B). SSCrFCES in medium represented>90% of the total recombinant protein expression. When stable cell line was cultured in serum-free medium without hygromycin for a week in a 1 L-Bellco spinner flask, a typical yield ˜1.6 mg/L of SSCrFCES was achieved.

[0097] The presence of SSCrFCES in the culture medium thus contributes to the ease of batch-continuous culture and purification. Most significantly, SSCrFCES expressed and secreted from insect cells was biologically active.

EXAMPLE 2

ELISA-Based Lipid A Binding Assay

[0098] A Polysorp™ 96-well plate (Nunc) was first coated with 100 μl per well of various concentrations of lipid A diluted in pyrogen-free PBS. The plate was sealed and allowed to incubate overnight at room temperature. The wells were aspirated and washed 6 times with 200 μl per wash solution (PBS containing 0.01% Tween-20 and 0.01% thimerosal). Blocking of unoccupied sites was achieved using wash solution containing 0.2% BSA for 1 hour at room temperature. Subsequently, blocking solution was removed and the wells washed as described above. Varying concentrations of SSCrFCES were allowed to interact with bound lipid A at room temperature for 2 hours.

[0099] Bound SSCrFCES was detected by sequential incubation with rabbit anti-SSCrFCES antibody (1:1000 dilution) and goat anti-rabbit antibody conjugated with HRP (1:2000 dilution) (Dako). Incubation with each antibody was for 1 h at 37° C. with washing between incubations as described above. In the final step, 100 μl of peroxide substrate ABTS (Boehringer Mannheim) was added. Using a microtiter plate reader, the absorbance of the samples was determined at 405 nm with reference wavelength at 490 nm. The values were correlated to the amount of LPS bound and SSCrFCES present. Quantitation of SSCrFCES was achieved from a standard curve derived by immobilizing known amount of purified SSCrFCES onto a Maxisorp plate.

[0100] Results from the ELISA-based lipid A binding assay displayed a biphasic curve (FIG. 2A). Unlike other LPS-binding proteins (18-21), SSCrFCES has multiple binding sites for the ligand. SSCrFCES binds co-operatively to lipid A with a stoichiometry of one SSCrFCES to ˜3 lipid A molecules at saturation (FIG. 2B). Scatchard plots of the binding data are very convex, indicating that the binding of SSCrFCES to lipid A is highly cooperative, being comparable to haemoglobin for oxygen (22). This is confirmed by the slope of the line obtained from plotting the data (FIG. 2C) according to the Hill's equation (23), which gave a coefficient of 2.2. While bactericidal/permeability-increasing protein (BPI) (18) was reported to bind>1 lipid A molecule, it was not reported to exhibit cooperativity in binding. This homotropic cooperativity for binding to lipid A is thus novel and unique to SSCrFCES.

[0101] The presence of multiple lipid A binding sites that showed cooperativity assuredly confirm the LPS-binding domain of Factor C, as well as full-length Factor C, to be the best candidate for removal and detection of endotoxin in solution, and supports its use as an anti-endotoxin therapeutic. Cooperative binding also contributed to Factor C's ability to detect sub-picogram level of endotoxin (U.S. patent application Ser. No. 09/081,767) as well as a competitive binding advantage over Limulus Anti-LPS Binding Factor (LALF).

[0102] Retrospectively, the degranulation of amoebocytes in the presence of LPS would release a battery of anti-bacterial/LPS binding factors e.g. LALF, thus significantly reducing the amount of free LPS. Nonetheless, Factor C is capable of capturing trace LPS to activate the coagulation cascade. Such capability is attributed to its homotropic cooperativity as demonstrated by SSCrFCES, that is to say, its LPS-binding domain.

EXAMPLE 3

Surface Plasmon Resonance (SPR) Studies on Biospecific Binding Kinetics Between Lipid A and: CrFCES; SSCrFCsushi-1,2,3-GFP: SSCrFCsushi-1-GFP; SSCrFCsushi-3-GFP: and synthetic Peptides

[0103] Recognition of lipid A by the abovenamed secreted recombinant proteins and peptides was, performed with a BIAcore X™ biosensor instrument and an HPA sensor chip. Briefly, lipid A at 0.5 mg/ml in PBS was immobilized to a HPA sensor chip (Pharmacia) according to the manufacturer's specification. In all experiments, pyrogen-free PBS was used as the running buffer at a flow rate of 10 μl/min.

[0104] With purified SSCrFCES, 4 μg/ml was injected into the flow cell at a rate of 10 μl/min, and the binding response was measured as a function of time. Following injection of SSCrFCES, a solution of INDIA™ HisProbe™-HRP antibody, diluted in PBS to 400 μg/ml, was also injected to cause a shift in SPR in order to further confirm that SSCrFCES binds to lipid A. For regeneration, 100 mM of NaOH solution was injected for 5 minutes. Similar lipid A binding analysis was carried out with SSCrFCsushi-GFP fusion proteins.

[0105] FIG. 3A shows that injection of 400 ng/100 ul of SSCrFCES over immobilized lipid A resulted in an increase of 200 relative response unit. This represents a 92% saturation of lipid A. Subsequently, injection of antibody (INDIA™ His-HRP Ab) against the poly-His tag of SSCrFCES resulted in a further increase of relative response unit. The binding of INDIA™ His-HRP Ab further confirms that only SSCrFCES was bound to the immobilized lipid A.

[0106] FIGS. 3B, 3C, and 3D show SPR (in response units) of the realtime binding interactions between SSCrFCsushi-1,2,3, SSCrFCsushi-1, and SSCrFCsushi-3-GFP fusion proteins, respectively, to the immobilized lipid A on the biochip. FIG. 3E shows the same binding interaction analysis of four examples of synthetic peptides derived from sushi-1 and sushi-3 of Factor C.

EXAMPLE 4

SSCrFCES and Synthetic Peptides are Potent Anti-Endotoxin Molecules—(Inhibition of Endotoxin-Induced LAL Reaction)

[0107] The Limulus Kinetic-QCL is a quantitative, kinetic assay for the detection of gram-negative bacterial endotoxin. This assay utilizes the initial part of LAL endotoxin reaction to activate an enzyme, which in turn releases p-nitroaniline from a synthetic substrate, producing a yellow color. The time required before the appearance of a yellow color is inversely proportional to the amount of endotoxin present. Throughout the assay, the absorbance at 405 nm of each well of the microplate was monitored. Using the initial absorbance reading of each well as its own blank, the time required for the absorbance to increase 0.200 absorbance units were calculated as Reaction Time. The 50% endotoxin-neutralizlng concentration (ENC50) reflects the potency of SSCrFCES or the synthetic peptides; a low ENC50 indicates high anti-endotoxin potency.

[0108] Briefly, 25 μl of endotoxin solution (LPS, E. coli 055:B5) at 200 EU/ml was mixed with an equal volume of SSCrFCES at 1 μM, in a series of 2-fold dilutions in LAL reagent water in disposable endotoxin-free glass dilution tubes (BioWhittaker) and incubated at 37° C. for one hour. The reaction mixtures were each diluted 1000-fold with LAL reagent water. The endotoxin activity was then quantified with Limulus Kinetic-QCL. One hundred μl of the diluted test mixture was carefully dispensed into the appropriate wells of an endotoxin-free microtitre plate (Costar). The plate was then pre-incubated for >10 minutes in a temperature-controlled ELISA plate reader. Near the end of the pre-incubation period, 100 μl of freshly reconstituted Kinetic-QCL reagent was dispensed into the wells using an 8-channel multipipettor. The absorbance at 405 nm of each well of the microtitre plate was monitored at time intervals of 5 minutes over a period of 2 hours. A 5 second automix was activated prior to reading. In the Limulus Kinetic-QCL, the assay was activated by 0.005 EU/ml of endotoxin.

[0109] The high sensitivity of the assay allowed for very low levels of endotoxin to be detected. Following incubation of endotoxin with SSCrFCES, a 1000-fold dilution was introduced to eliminate any potential effects of the SSCrFCES on the LAL enzyme system. A sigmoidal curve is usually expected between relative reaction time and the logarithmic concentration of the SSCrFCES. The best fit curve was derived using SigmaPlot and the concentration corresponding to 50% relative increase in reaction time was designated ENC50. The mean values were obtained from 3 independent experiments.

[0110] The time that is required before the appearance of a yellow color (Reaction Time) is inversely proportional to the amount of endotoxin present. A low ENC50 indicates high potency of endotoxin neutralization. The ENC50 is taken as the concentration of SSCrFCES that reduces the mean reaction time by 50%. A sigmoidal curve was obtained between relative reaction time and the logarithmic concentration of SSCrFCES (FIG. 4). ENC50 of SSCrFCES was determined to be 0.069±0.014 μM. Comparatively, this value is 28- and 7.5-fold less than ENC50 of polymyxin B and LF-33 (33-mer peptide derived from lactoferrin) (24), respectively. This shows that on a molar basis, much less SSCrFCES is required to neutralize the same amount of LPS. Consequently, it also indicates that SSCrFCES is a potent anti-pyrogenic recombinant protein.

[0111] The ENC50 of the synthetic sushi peptides were comparable to other reported peptides, e.g.: S1=2.25 μM; S1Δ=0.875 μM; S3=1 μM; A3=0.625 μM. For the designed peptides, the ENC50 values were: V1=0.47 μM and V2=0.89 μM.

[0112] Hill's plot for the interaction between synthetic peptides and lipid A shows that S1 exhibited high positive co-operativity of n=2.42, indicating that more than two S1 peptides interact with one LPS molecule.

EXAMPLE 5

The Anti-Sepsis Activities of SSCrFCES and Synthetic Peptides: Inhibition of the LPS-induced TNF-α and IL-8 by (a) THP-1 Cells (b) Human Peripheral Blood Mononuclear Cells (PBMC)

[0113] During gram-negative bacterial septicaemia, the high concentration of LPS in the blood leads to multiple organ failure syndromes. These adverse effects are dependent on the generation of endogenous mediators. A multitude of mediators have been implicated, including arachidonic acid metabolites, PAF, cytokines such as TNF-α; interferons, and various interleukins (e.g. IL-1, IL-8, etc.), reactive oxygen metabolites, and components of the coagulation cascade (1-3). Consequently, the biological potential of SSCrFCES to bind and neutralize LPS-stimulated production of cytokines in human promonomyelocytic cell line THP-1 and normal human PBMC were investigated.

[0114] Results from our in vitro binding studies suggested that SSCrFCES would be a potent inhibitor of the LPS activation of monocytes. To test this prediction, we measured the ability of SSCrFCES to inhibit hTNF-α and hIL-8 production by THP-1 cells incubated with 25 ng/ml and 100 ng/ml of LPS in a serum-free system containing various concentrations of SSCrFCES. THP-1 cells were grown in RPMI 1640 medium supplemented with 10% FBS, penicillin (100 U/ml) and streptomycin (0.1 mg/ml), at 37° C. in a humidified environment in the presence of 5% CO2. The cells were maintained at a density between 2.5×105 and 2.5×106 cells/ml.

[0115] THP-1 cells were prepared for experiment by addition of a concentrated stock solution of phorbol myristate acetate (PMA, 0.3. mg/ml in dimethyl sulfoxide) to cell suspension to give a final concentration of 30 ng/ml PMA and 0.01% dimethyl sulfoxide (25). PMA-treated cell suspensions were immediately plated into 96-well microtitre plate at a density of 4×105 cells/ml and allowed to differentiate for 48 hours at 37° C. Immediately before stimulation by 25 ng/ml LPS or LPS pre-incubated with various concentrations of SSCrFCES, the culture medium was removed, and the cells were washed twice with serum-free RPMI 1640 and incubated at 37° C. At indicated times, the culture medium was collected. Human TNF-α and IL-8 concentrations in the supernatants were assayed using ELISA as suggested by the manufacturer.

[0116] Heparinised venous blood drawn from healthy donors was subjected to fractionation using Ficoll-Paque PWS (Pharmacia) to obtain peripheral blood mononuclear cells (PBMC). PBMC were washed with PBS and suspended at a cell density of 1.5×106 cell/ml with RPMI 1640 medium supplemented with 10% FBS. PBMC were incubated at 37° C. for 24 h at a density of 1.5×105 per well. LPS stimulation and immunoassay of hTNF-α and hIL-8 were performed as described for THP-1 cells. In addition, the suppressive effect of SSCrFCES on LPS-induced cytokine release was investigated in the presence of 10% human serum. The difference between the test and control groups was subjected to Student's t-test. The values were obtained from at least three independent experiments.

[0117] FIG. 5 shows that with THP-1 cells, 0.5 μM of SSCrFCES potently inhibited >90% LPS-induced production of TNF-α and IL-8 in the presence of high level of endotoxin. At 25 ng/ml LPS concentration tested, 0.7 μM of SSCrFCES is sufficient to completely prevent LPS-induced TNF-α production (FIG. 5A). At 10 ng/ml LPS, 1 μM of SSCrFCES reduced 90% IL-8 production as compared to control (FIG. 5B).

[0118] Our findings indicate that 1 μM of SSCrFCES effectively prevent the LPS-mediated induction of hTNF-α and hIL-8 production by THP-1 when these cells were incubated in the presence of high endotoxin levels. It is important to note that the concentrations of LPS (25 ng/ml and 100 ng/ml) used in these studies are among the highest known concentrations reported for LPS-induced cytokine production. On molar basis, SSCrFCES appears to be more potent than polymyxin B and LF-33 at suppressing LPS-induced LAL coagulation and hTNF-α or hIL-8 secretion by THP-1 cells under serum-free conditions (24). This suggests that SSCrFCES has a much greater intrinsic capacity to neutralize endotoxin than polymyxin B. Again, it is attributable to its cooperative binding of LPS.

[0119] Purified human PBMC were used to test the suppression of endotoxin-induced TNF-α and IL-8 secretion by SSCrFCES under normal physiological conditions. In the absence of human serum, addition of only 0.1 μM of SSCrFCES completely inhibited TNF-α and IL-8 response to 10 ng/ml LPS by 50% (FIGS. 6A and 6B). When SSCrFCES was added to human serum (final concentration, 10%) before the addition of endotoxin, the suppressive effect of SSCrFCES was attenuated. It required 17 fold more SSCrFCES to suppress TNF-α and IL-8 secretion by 50%. A similar effect of human serum has also been observed with other cationic anti-endotoxin proteins such as LF-33 (24) and LALF (26). This is due to the interaction of these factors with serum proteins that effectively reduce their availability for binding to endotoxin. However, if the SSCrFCES was mixed with endotoxin 5 min before the addition of serum, the effect of the serum on the neutralization of endotoxin by SSCrFCES was greatly reduced, requiring only 4 fold more SSCrFCES for 50% inhibition (FIGS. 6A and 6B).

[0120] Results from the in vitro binding studies suggested that the 4 Factor C-based sushi peptides would be potent inhibitors of the LPS-induced cytokine release by monocytes. To test this prediction, we measured the ability of S1, S1Δ, S3, and S3Δ to inhibit hTNF-α production by THP-1 cells incubated with 10 ng/ml of LPS in a serum-free system containing various concentrations of peptides.

[0121] As shown in FIG. 6C, both modified peptides, S1Δ and S3Δ, are more potent inhibitors, giving 50% inhibition at 53.3 and 45.8 μM, respectively, as compared to the S1 and S3 peptides.

[0122] With the designed peptides (V1 and V2) 50% inhibition of LPS-induced TNF-α release were 27 and 35 μM, respectively.

EXAMPLE 6

SSCrFCES and Synthetic Peptides are not Cytotoxic to Eukaryotic Cells

[0123] In addition to high specific LPS binding, an important feature when using proteins for in vivo application to treat Gram-negative bacterial septic shock, are their physicochemical properties in biological systems. Problems that often arise in animal experiments are due to toxicity, as in the case of polymyxin B, or a very short half-life in the circulating system, for example BPI. To assess these features, we investigated SSCrFCES for their ability to permeabilize cultured cells.

[0124] Two×104 THP-1 monocytes in 50 μl of RPMI 1640 were mixed in a microtitre plate with 50 μl of increasing amount of 2-fold serial dilutions of SSCrFCES (0.004-4.0 mg/ml in PBS) and incubated for 60 min at 37° C. To determine cytotoxicity induced by the SSCrFCES, 20 μl of CellTiter96™ AQueous One Solution Reagent (Promega) was added into each well for 90 min at 37° C. [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) is bioreduced by metabolically active cells into a colored formazan product that is soluble in tissue culture medium (27, 28). For detection, the absorbance was measured at 490 nm. To determine the ratio of cell lysis induced by SSCrFCES, two controls were used. Complete lysis (100%) was achieved by incubating cells in phosphate buffer saline containing 0.2% Tween-20 instead of medium only. This absorbance value corresponded to the background, as those cells could not metabolize MTS. The second control representing 0% lysis was determined by incubating cells in medium only. The LD50 was calculated as the concentration of SSCrFCES necessary to lyse 50% of the cells. The experiment was done in triplicate.

[0125] SSCrFCES had minimal effect on cell permeabilization (FIG. 7). At the highest concentration of 4 mg/ml or 109 μM, only 20% cell lysis was observed. Compared to polymixin B where 50% cell lysis occurred with 0.51 mg/ml (29), this clearly indicates that SSCrFCES is a non-toxic anti-endotoxin protein. The synthetic peptides are non-cytotoxic.

EXAMPLE 7

Pharmacokinetic Analysis of SSCrFCES—Clearance Rate in Mice

[0126] 600 μg of SSCrFCES was labeled with biotin using EZ-Link™ PEO-Maleimide Activated Biotin (Pierce) according to the manufacturer's instructions. The excess biotin was subsequently removed via ultrafiltration through Microcon-10 (Millipore). Three C57BL/6J mice were given a single i.v. bolus injection of 200 μg biotin-labeled SSCrFCES. Blood was collected from each of the 3 mice at time intervals over a 5-h period. The blood was immediately treated with SDS-PAGE loading dye and boiled for 5 minutes. The mixture was resolved in a 12% SDS-PAGE and electroblotted onto a PVDF membrane. Immunoblotting and hybridisation were carried out as described above except NeutrAvidin™-HRP antibody (Pierce) was used. Exposure time for chemiluminescence detection was extended to 1 hour. The signal on the X-ray film was quantitated via densitometric scan. The clearance rate of biotin-labeled SSCrFCES was analyzed using NCOMP, which is a WINDOWS-based program for noncompartmental analysis of pharmacokinetic data (30).

[0127] Densitometric scan revealed that significant amounts of circulating half-life of SSCrFCES is sufficiently long to allow easy detection during the first 90 minutes post-injection. NCOMP, which provides an interactive graphical environment for noncompartmental analysis of pharmacokinteics data by facilitating estimation of the zero and first moments of concentration-time data, was used for analysis. The calculated clearance rate of biotin-labeled SSCrFCES in C57BL/6J mice is 4.7 ml/min (FIG. 8). The clearance rate is 2.7 fold slower than BPI. Therefore, a lower dose of SSCrFCES would be adequate to maintain high enough circulating levels to compete with LBP for LPS.

EXAMPLE 8

SSCrFCES and Synthetic Peptides Neutralize LPS-Induced Lethality in Mice

[0128] The anti-endotoxin potency of SSCrFCES was investigated in C57BL/6J mice. Mice are typically resistant to endotoxin. However, the sensitivity of mice to endotoxin can be enhanced>1,000-fold by co-injection with a liver-specific inhibitor, galactosamine (31). In our study, intraperitoneal (i.p.) injection of 2.5 ng of E. coli 055:B5 LPS together with 15 mg of galactosamine hydrochloride in 0.2 ml of saline induced nearly 100% lethality in 18-25 g C57BL/6J mice within 7 hours. Various concentrations of SSCrFCES (1, 2, and 4 μM) and synthetic peptides (25 and 75 μg) were injected intravenously (i.v.) through tail vein 10 minutes after i.p. injection of the LPS-galactosamine mixture. Lethality was observed over 3 days after injection. Statistical analysis were performed using the Kaplan-Meier test (32) and log rank pairwise test.

[0129] As shown in FIG. 9A, the LPS-induced lethality was reduced by 20% when 1 μM of SSCrFCES was injected i.v. 10 min after the i.p. injection of LPS. Higher concentrations of SSCrFCES of 2 and 4 μM conferred 90% and 100% protection, respectively.

[0130] A protective role of SSCrFCES viz LPS-binding domain of Factor C is thus shown in an intraperitoneal murine sepsis model. The mechanism by which SSCrFCES protects mice from LPS-induced sepsis is presumably mediated through its high affinity association to lipid A moiety of LPS, which consequently reduces the secretion of cytokines like TNF-α and IL-8. FIG. 9B shows that S1, S1Δ, and S3 conferred 22-100% protection, whereas at 75 μg, S3Δ was most efficacious, giving 100% protection against LPS-induced toxicity.

EXAMPLE 9

Antimicrobial Action

[0131] Recently, the concept of eradication via targeted disruption of bacterial LPS by cationic peptides/proteins was introduced (33). For an effective antimicrobial therapy, such peptides need to satisfy several important criteria, including potent antimicrobial activity over a wide range of pH, fast killing rate, low toxicity, and low hemolytic activity. While numerous antimicrobial peptides/proteins, like FALL-39 (34), SMAP-29 (35), lepidopteran cecropin (36), and CAP-18 (37) have been reported, few display all the above mentioned attributes. Thus, the search for new, more powerful and yet safe antimicrobial peptides continues to enrich the therapeutic armamentarium.

[0132] Further analysis of the sushi peptides showed them to have low cytotoxicity and to be capable of neutralizing LPS-biotoxicity (See Examples 4, 5, and 6 above). This property provides a vital advantage over other antimicrobial peptides in suppressing adverse effects of LPS-induced septic shock during or after treatment.

[0133] Septic shock is characterized by a drastic fall in blood pressure, cardiovascular collapse, and multiple organ failure. Septic shock is responsible for over 100,000 deaths a year in the US alone. The septic shock condition (38, 39) often creates more complication than the actual infection itself when a massive amount of LPS is released by bacteria disintegrated by antibiotics. This problem is especially pronounced in children, in the elderly, and in immuno-compromised patients.

[0134] The present invention demonstrates novel and hitherto unsurpassed antimicrobial action of Factor C sushi peptides against clinical isolates of P. aeruginosa. Although the sushi peptides are demonstrated to be efficacious against this microorganism, antimicrobial potency is not limited to P. aeruginosa but should extend to any bacterium producing LPS bound by Factor C.

[0135] Antimicrobial action of SSCrFCES, SSCrFCsushi-GFP proteins and synthetic peptides (e.g., S1, S1Δ, S3, S3Δ, and V peptides ), examined by microbiocidal concentrations (MBC90) assays, show that these recombinant proteins and synthetic peptides have potent antimicrobial activities. Antimicrobial activity is expressly demonstrated against Pseudomonas aeruginosa and Klebsiella pneumoniae and Helicobacter pylori. However, the antimicrobial activity of these proteins and peptides is not limited to only these three species of bacteria.

[0136] Peptides of 34 amino acids were synthesized based on the sequence of two regions of Factor C: sushi 1 and sushi 3, as well as their corresponding mutants (sushi 1Δ and sushi 3Δ), were found to harbour strong antimicrobial activities. Collectively, all four peptides (named S1, S1Δ, S3, and S3Δ) demonstrated exceptionally effective bactericidal activity against gram-negative bacteria, represented by Pseudomonas aeruginosa.

[0137] At 03-0.25 μg/ml (8-63 nM), the MBC90 values of the peptides, are of the lowest ever reported against Pseudomonads. Viable bacteria were reduced by 90% after 7 minutes and were totally eradicated within 30-40 minutes. These peptides were minimally hemolytic against both rabbit and human erythrocytes (30%) at concentrations of 100 μg/ml (25 μM), which is up to 3333 times their effective MBC concentration.

[0138] These findings demonstrate the unprecedented therapeutic value of the sushi peptides and their mutants for treatment of Pseudomonas infections. Other sushi peptide derivatives (S4, S5) were also found to have variable antimicrobial activities. Thus, these results are given by way of example, and the present invention should not be deemed to be limited to only these representative peptides.

[0139] Test strains cultured on Mueller-Hinton agar (MHA, Becton Dickinson, USA) were inoculated into 10 ml Mueller-Hinton broth (MHB, Becton Dickinson, USA) and grown overnight at 37° C. in a shaker incubator (Model 4536, Forma Scientific, Inc., USA) at 230 rpm. Overnight broth cultures were diluted to give a final cell density of 105 colony forming units/ml (cfu/ml). One hundred microliters of the bacterial suspension was dispensed into sterile polypropylene 8-strip PCR-tubes (Quality Scientific Plastics, USA). Eleven microliters of serially diluted sushi peptides, ranging in final concentrations of 0.03-4 μg/ml, were then added. The peptides were constituted at 10 times the required test concentrations in 0.01% acetic acid and 0.2% bovine serum albumin (BSA). Positive controls were cultures without test peptides. Uninoculated MHB was used as negative control. All tests were carried out in triplicate.

[0140] Cultures were incubated at 37° C. for 18-24 h, with the PCR-tubes held in horizontal position and shaken at 230 rpm. Cell counts were determined by standard drop-count method. The killing efficiency for the four sushi peptides were calculated based on standard drop-count method. All four peptides (S1, SΔ1, S3, and SΔ3) showed potent bactericidal activity of <0.03-0.25 μg/ml against the 30 clinical strains of P. aeruginosa (Table 2).

[0141] The MBC50 determined for all the 4 peptides was<0.03 μg/ml (<7.5-8.0 nM). The MBC90 for the peptides were: <0.03 μg/ml (8 nM) for S1; 0.06 μg/ml (16 nM) for SΔ1; <0.03 μg/ml (8 nM) for S3; and 0.25 μg/ml (63 nM) for SΔ3. These MBC90 values are unsurpassed by any known antimicrobial peptides reported for P. aeruginosa. The MBC90 for the control strain of P. aeruginosa ATCC 27853 was 0.03 μg/ml (7.5-8.0 nM) for all the 4 peptides (FIG. 12).

[0142] The antimicrobial therapeutic value of sushi peptides is exhibited by their exceptional bactericidal activity against gram-negative bacteria, e.g.: 30 clinical isolates and a control strain of P. aeruginosa ATCC 27853. The resistance pattern of these strains gave a close representation of the resistant strains of P. aeruginosa found in Singapore (Table 2).

[0143] The remarkably low MBC90 values of <0.03-0.25 μg/ml (<8.0-63 nM) obtained for the peptides are unsurpassed by any known antibiotics of metabolite or peptide origin. Comparatively, sushi peptides are 1-3 orders of magnitude more effective against P. aeruginosa than are other reported antimicrobial peptides. Owing to their high affinity for LPS, the sushi peptides probably exert anti-Pseudomonas effect through disruption of the LPS-lamellar organization.

[0144] Although, the peptides are targeted at the conserved lipid A domain, different MBCs were observed over the 30 clinical isolates. This is most likely due to differential permeability of the peptides into the variable polysaccharide components in the different Pseudomonas strains. This is supported by the different binding affinities of the sushi peptides for Escherichia coli B5:055 lipid A (See Example 3 of the present application).

[0145] The killing rate assay was adapted from the MBC test above, with different contact time of peptides with the bacteria arrested at regular intervals and plated for colony count. An initial density of 109 cfu/ml was used. FIG. 13 shows that sushi peptides exhibit rapid bactericidal action. This is one of the important features of an effective therapeutic agent.

[0146] With an effectively low MBC90 concentration, we proceeded to investigate the killing time for the sushi peptides. At 0.06 μg/ml, all four peptides achieved MBC90 within 7 minutes. Within 30 minutes, the peptides totally eradicated an initial cell population of 1×109 cfu/ml (FIGS. 13 and 14). P. aeruginosa is a fast-replicating bacteria, which displays a short lag phase and doubling time. Hence, a rapid bactericidal action is an extremely important factor especially with an infection that occur near or in vital organs like cornea (contact lens contamination in the eye), lung (in cystic fibrosis), and acute bacteraemia in AIDS patients. At a concentration of 0.06 ug/ml, the sushi peptides were able to eradicate 90% of viable cells within 7 min of incubation (FIG. 13).

[0147] Complete eradication is assured to occur within the first two generations of bacteria which reduces the possibility of mutation. Thus, this rapid killing rate reduces the chance/opportunity for the development of resistance. Resistance will be remote as it will require several precise mutations occurring at multiple enzymes along the LPS synthesis pathway to ultimately yield a modified LPS structure that is sufficiently different to evade sushi peptide recognition. However, the possibility of developed or acquired resistance cannot be precluded if some of these strains are allowed to mutate at sub-lethal peptide concentrations.

[0148] FIG. 15 shows electron micrographs illustrating how some multiple antibiotic-resistant strains of bacteria are killed by these peptides. Human and rabbit erythrocytes were both used to test the hemolytic activities of the peptides. Whole blood was collected in heparinized sterile syringe, transferred to sterile borosilicate tube and centrifuged at 1200 g for 5 minutes at 4° C. The supernatant including the leukocytes above the erythrocyte pellet was discarded. The erythrocytes were washed 3 times using three volumes of prechilled pyrogen-free saline (PFS). An erythrocyte suspension at 0.4% was prepared for the hemolysis assay. Serial two-fold dilutions of the peptides was prepared in PFS and 100 μl aliquots were added to equal volumes of 0.4% erythrocyte solution in a 96-well microtiter plate (Nunclon™ Δ surface, Nunc) to give final peptide concentrations ranging from 6 to 100 μg/ml. The mixtures were incubated at 37° C. for 1 h. The intact erythrocytes were then pelleted by centrifuging at 1000 g for 5 min. One hundred μl of the supernatant was transferred to a new 96-well microtiter plate and the amount of hemoglobin released into the supernatant was determined by reading the absorbance at 414 nm using a SPECTRAmax™ 340 plate reader with SOFTmax PRO™ version 1.2.0. A positive control with 100 μl of 0.4% erythrocyte lysed in 1% Triton-X 100 was taken as 100% lysis. The negative control was the erythrocytes in PFS alone, which gave minimal lysis. This was taken as 0%.

[0149] FIG. 16 shows that sushi peptides have low hemolytic activity. This is crucial to the applicability of an antimicrobial agent for therapeutic use in humans and animals. Even at concentrations of 100 μg/ml (25 μM), up to 400-3333 fold of their MBC90, the sushi peptides showed minimal hemolytic activity (FIG. 16). On a separate assay, the hemolytic activity of sushi peptides was tested on rabbit erythrocytes. At the same concentration, the peptides showed hemolytic activity below 6%. For purposes of the present application, the language “substantially free of hemolytic activity” means showing hemolytic activity below 6%.

[0150] Thus, the ability of sushi peptides to: (a) cause effective LPS-neutralization (see Examples 4 and 5); (b) confer crucial protection against LPS-indcued lethanlity in mice (see Example 8); (c) possess low MBC90 values; (d) induce rapid killing rate; and (e) exhibit lack of hemolytic activity, are features that indicate that these peptides will provide a great advantages over currently available antibiotics.

[0151] With this invention, the LPS toxicity during the course of treatment will be dramatically reduced. The sushi peptides will provide highly effective and potentially useful therapeutics for the treatment of P. aeruginosa infections. It leaves very little doubt that these peptides will be equally effective against other members of Pseudomonads.

EXAMPLE 10

SSCrFCsushi-GFP Proteins Bind LPS and Gram-Negative Bacteria

[0152] The recombinant SSCrFCESsushi-GFP proteins were able to bind/tag gram-negative bacteria, showing as green fluorescent tagged organisms. This makes a convenient detection tag for displaying such microogranisms in samples.

EXAMPLE 11

LPS-Affinity Chromatography (for Removal of Endotoxin from Liquid Samples)

[0153] By way of an example, S3Δ peptide (with Kd of 10−7 to −8 M) was chosen from amongst the sushi peptides to create an affinity chromatography system to display the power of binding of LPS from liquid samples. Thus, a solution of 4 mg/ml of S3Δ (in conjugation buffer: 0.1 M MES [2-(N-Morpholino)ethanesulfonic acid], 0.9% NaCl, pH 4.7) was immobilized via EDC [1-ethyl-3-(3-dimethylaminopropyl)cardiimide]/DADPA (Diaminodipropylamine), obtained from Pierce Chemicals, USA). After 3 hours of conjugation to DADPA-Agarose CL-6B in a small column, the flowthrough was collected and the absorption of fractions at 280 nm was measured to calculate the total amount of peptide immobilized to the matrix (by substraction from the unbound S3Δ found in the flowthrough).

[0154] It was found that binding efficiency of S3Δ to the EDC-activated resin was 50%. After regeneration of the column with 5 column volumes of 1% sodium deocycholate (DOC)—to ensure the removal of any exogenous LPS that may be bound to the resin, and washing the resin with pyrogen-free water, the column was ready for LPS absorption.

[0155] Again, by way of example, two 50 ml volumes of LPS solution (either LPS from Sigma, or FITC-labelled LPS from List Biologicals) containing 1 and 0.05 EU/ml were loaded onto the column. In each case, the flowthrough was subjected to LPS measurement by either LAL kinetic-QCL kit (BioWhittaker) or spectrofluorimetry, depending of the type of LPS solution that was used. In each case, the level of unbound LPS remaining in the flowthrough was below the detection limit (0.005 EU/ml) of the LAL kinetic-QCL assay. The affinity column was re-usable repeatedly, using 1% DOC as a regenerating agent.

[0156] LPS-affinity chromatography was also demonstrated by batchwise chromatography using 0.5 ml of 0.5 ug/ml FITC-LPS solution (in different buffers). The resin suspension was rotated for 3 h at room temperature, briefly spun at 1000 rpm for 1 min and the supernatant was reclarified at 12000 rpm for 10 minutes. The resultant supernatant was measured for unbound FITC-LPS by spectrofluorimetry. FIG. 17 shows S3Δ peptide-FITC-LPS coupled agarose beads seen under UV-fluorescence microscope.

[0157] The optimal binding of LPS to S3Δ was tested under different pH conditions and ionic strengths. Binding decreases with increase in ionic strength, and increases with increase in pH (FIG. 18). Thus, the best condition for binding of LPS to the affinity resin is basic and low ionic strength conditions. The optimal condition is expected to vary with different protein solutions.

[0158] Purified SSCrFCsushi-GFP proteins can also be chemically-linked to activated resins via their C-terminus GFP region, to allow N-terminal LPS-binding domain to be exposed for capturing endotoxin when an LPS-contaminated solution or biological fluid is passed through the resin.

EXAMPLE 12

Production of rFC in P. pastoris

[0159] In this study, the cloned Factor C cDNA of the Singapore horseshoe crab, Carcinoscorpius rotundicauda (10), was expressed in a methylotrophic yeast, Pichia pastoris. The full-length rFC so produced was found to lack serine protease activity, yet possess a functional endotoxin-binding domain. The full-length rFC from P. pastoris is able to bind free or bound LPS. Deletion proteins rFCEE and rFCSN containing the 5′ and 3′ regions, respectively, of Factor C were also produced and assayed for lipid A binding activity. The presence of a fully functional endotoxin-binding domain on the full-length rFC, and a slightly reduced endotoxin-binding capacity in rFCEE was demonstrated by two modified qualitative and quantitative LPS binding assays.

[0160] A. Materials and Methods

[0161] (1) Glassware and Buffers

[0162] All glassware was rendered pyrogen-free by baking at 200° C. for 4 h. Buffers were prepared using pyrogen-free water (Baxter) and autoclaved at 121° C. for 2 h. Sterile disposable plasticware was used whenever possible. Other non-heat resistant apparatus was soaked in 5% hydrogen peroxide before rinsing with pyrogen-free water and drying in an oven.

[0163] (2) Recombinant Factor C Constructs

[0164] Three recombinant Factor C constructs—pHILD2CrFC21, pHILD2/CrFC21EE, and pPIC9/CrFC26SN (10,12,75)—were used for the study. As a control, pHILD2/151, an isolate of P. pastoris containing only the parent vector, pHILD2, was also included. pHILD2/CrFC21 contains the full-length CrFC21 cDNA (GenBank Database Accession No. S77063) of 3.4 kb together with its native translational start and signal sequence while pHILD2/CrFC21EE contains the 2.3 kb 5′ EcoRI fragment isolated from CrFC21 cDNA. This construct contains the 762 amino-acid fragment encompassing the heavy chain of CrFC21 along with its endotoxin-binding domain. The pPIC9/CrFC26SN construct contains the 2.4 kb 3′ SalI-NotI fragment of CrFC26 (GenBank Database Accession No. S77064) cloned as a fusion fragment, in-frame and downstream of the pPIC9 vector start site and secretion signal. CrFC26SN contains sequence similar to the corresponding fragment in CrFC21 (10). This is a truncated construct lacking the putative LPS-binding domain and therefore, serves as a useful negative control in LPS-binding assays. The recombinant Factor C proteins from pHILD2/CrFC21, pHILD2/CrFC21EE, and pPIC9/CrFC21SN are referred to as rFC, rFCEE, and rFCSN, respectively.

[0165] (3) Growth Conditions

[0166] Recombinant Pichia clones of pHILD2/CrFC21, pHILD2/CrFC21EE and pPIC9/CrFC26SN as well as the negative control, pHILD2/151 were grown overnight in shake flasks at 300 rpm and 30° C. in 1 L MGY growth medium containing 1.34% yeast nitrogen base (Difco), 1% glycerol and 4×10−5% biotin. At the mid-log phase of growth (OD600 2.0), the yeast cells were harvested aseptically at 3,000×g for 10 min and transferred to 2 L MM induction medium, containing 1.34% yeast nitrogen base (Difco), 0.5% methanol and 4×10−5% biotin. Induction was carried out at 30° C. for 8 h. Induced cells were harvested by centrifugation at 3,000×g for 10 min.

[0167] (4) Preparation of rFC Samples from Recombinant Yeast Clones

[0168] Induced yeast cells were disrupted by 10 cycles of nebulization (Glas-Col™ BioNeb) at 200 psi using purified N2. Soluble and insoluble fractions were separated by centrifugation at 13,200×g for 12 min. The supernatant containing soluble proteins was partially purified by ammonium sulfate precipitation at 20% saturation and resuspended in 50 mM Tris-Cl buffer, pH 8. The mixture was desalted through a Sephadex™ G-25 column (Pharmacia) equilibrated in the same buffer. In a separate preparation, the crude yeast supernatant was subjected to ultrafiltration through a Biomax™-50 (Millipore) membrane. The Biomax™-50 enriched rFC was further purified by chromatography through a Sephadex™G-100 column (1×25 cm; Pharmacia). Total protein was measured by Bradford assay (62).

[0169] (5) Western Analysis of rFC Protein

[0170] rFC samples were electrophoresed on denaturing 10% SDS-polyacrylamide gel (63) and electroblotted onto Immobilon™ PVDF membrane. The respective rFC was immunolocalized by incubating the blot with rabbit anti-Factor C primary antibody and visualizing with horseradish peroxidase-conjugated secondary goat anti-rabbit antibody (Dako) using 4-chloro-1-naphthol and H2O2 as substrate.

[0171] (6) LPS-Binding Assay of rFC

[0172] LPS from E. coli 055:B5 (Sigma) was reconstituted to 2 μg/μl, and diphosphoryl lipid A from E. coli K12, D31m4 LPS (List Biologicals, Inc., USA) was made up to 1 μg/μl. The LPS-binding assay was based on modifications of earlier described protocols (45,61). Briefly, 10 μg aliquots of LPS/lipid A were electrophoresed on a denaturing 15% SDS-polyacrylamide gel and electroblotted onto Immobilon™ PVDF membrane. The membrane was cut into strips and each LPS/lipid A strip was subsequently incubated with 300 μg of proteins containing rFC. Detection of rFC binding to lipid A was accomplished by incubation with anti-Factor C antibody followed by alkaline phosphatase-conjugated secondary goat anti-rabbit antibody (Dako) and BCIP/NBT colorimetric substrate (Moss, Inc., USA).

[0173] (7) Assay for Competition Between rFC and CAL Factor C for LPS

[0174] Carcinoscorpius amoebocyte lysate (CAL) containing native Factor C was used in an assay in which rFC competed with CAL Factor C for LPS. Because the rFC produced in P. pastoris lacks serine protease activity, the competition can be monitored by measuring the reduced enzymatic activity of CAL in a fluorimetric assay. Mixtures of 100 μl each of increasing concentrations of LPS or rFC, in fluorimetric assay buffer (50 mM Tris HCl pH 8, containing 0.1 M NaCl and 0.01 M CaCl2) were incubated at 37° C. for 1 h. Aliquots of 20 μg CAL were added to each mixture and the total volumes were made up to 2 ml with fluorimetric assay buffer. The reaction was continued at 37° C. for 1 h and the fluorimetric assay protocol (64) was followed. This involved the addition of 15 μl of 2 mM fluorimetric substrate N-t-Boc-Val-Pro-Arg-7-amido-4-methylcoumarin (Sigma) and incubation at 37° C. for 30 min. The reaction was terminated by the addition of 0.1 ml of glacial acetic acid (Merck). The product, amino methylcoumarin, was measured in fluorescence units (FU) on a Luminescence Spectrometer LS-5 (Perkin-Elmer) with excitation light at 380 nm and emission at 460 nm.

[0175] (8) Binding Interactions Between rFC and its Immobilized Ligand, Lipid A

[0176] The binding interactions between rFC and immobilized lipid A were monitored using the BIACORE X™ biosensor (Pharmacia Biotech). The BIACORE X™ sensor chip features a flat hydrophobic surface that allows the immobilization of ligand molecules. Thirty microliters of lipid A at 100 μg/ml were immobilized on each sensor chip to form a ligand surface. Biomax™-50 enriched samples of rFC, rFCEE and rFCSN, each at 1 mg/ml were injected at 10 μl/min for 3 min over the ligand surface. After each injection of the recombinant protein samples, the lipid A ligand surface was regenerated using 0.1 M NaOH. The ligand-binding was measured in relative response units (RU) for each sample, and calculated from the difference in RU at the baseline, viz., before injection of sample, and final experimental reading taken after sample injection and a 2-min wash. The percentage binding was thus determined.

[0177] B. Results and Discussion

[0178] Nebulization of P. pastoris Clones Released Soluble and Bioactive rFC.

[0179] After nebulization, the supernatant derived from clarification at 13,200 x/g of the P. pastoris cell lysate contained soluble forms of rFC and rFCSN of 132 and 89 kDa protein bands, respectively (FIG. 19). Compared to glass bead treatment (61), nebulization enhanced the breakage efficiency of P. pastoris. Furthermore, the rFC was fractionated into the soluble phase, thus enabling its direct use for functional analysis, as well as ease of purification. This is a significant improvement over the earlier rFC preparations from glass-bead breakage where insoluble rFC had to be solubilized by treatment with detergents (12). Detergent-solubilization, in particular, with Triton X-100 has been reported to inhibit Factor C binding of LPS (68). We have also shown that SDS at >0.5% also inhibits the activity of Factor C in CAL. Removal of SDS using potassium chloride (69) restores the LPS binding activity of solubilized rFC. However, care must be taken to avoid pyrogenation. Thus, it is best to obtain soluble rFC under pyrogen-free conditions via physical methods and not chemical means.

[0180] Using either (NH4)2SO4 precipitation or Biomax™-50 ultrafiltration, the rFC preparation was enriched in total protein content. Chromatography of Biomax™-50 rFC through a Sephadex™ G-100 molecular sieve further purified rFC from other yeast proteins (FIG. 20).

[0181] In the modified Western blot of LPS, rFC was shown to bind to the lipid A moiety displaying specific bands in the range of 7-20 kDa (FIG. 21A) which is consistent with previous findings (43,70). Subsequently, when the modified Western blot of diphosphoryl lipid A was used, the specificity of rFC for lipid A was further confirmed (FIG. 21B). Recombinant Factor C samples derived from (NH4)2SO4 precipitation; Biomax™-50 ultrafiltration and Sephadex™ G-100 gel filtration displayed increasing affinity for lipid A (FIGS. 21A and 21B: Lanes 3 & 4). No binding to the 7-20 kDa bands was observed with rFCSN and pHILD2/151. Biomax™-50 enriched rFCEE, the truncated Factor C, was also able to bind specifically to lipid A moiety of LPS, albeit less strongly (FIGS. 21A and 21B).

[0182] The presence of a functional LPS-binding domain demonstrates that rFC expressed in yeast folds properly, or at least its endotoxin-binding domain does so. The postulated endotoxin-binding region of the C rotundicauda Factor C (61) is located in the amino terminus of the heavy chain, which comprises the cysteine-rich EGF-like domain and one or two sushi domains (16,71). Recombinant Factor C specifically binds to the lipid A moiety showing the ability of rFC to recognize and bind the biologically-potent moiety of LPS. That the binding of LPS to Factor C requires the presence of an endotoxin-binding domain was confirmed using the rFCSN where the lack of 5′-terminal LPS-binding domain in this deletion homolog resulted in its inability to bind lipid A. The observable reduction in intensity of binding of rFCEE to lipid A as compared to that of rFC indicates that although the binding of lipid A requires the presence of the LPS-binding domain, sequences of Factor C further downstream may mediate the strength of the binding. It has been reported that binding of endotoxin triggers a conformational change in the Factor C molecule (58) where downstream sushi domains are involved in protein-protein interaction.

[0183] Recombinant Factor C competes for LPS, causing reduction in enzymatic activity of CAL Factor C. The yeast rFC, lacking serine protease activity, but capable of binding to LPS, was observed to compete with CAL native Factor C for the LPS. This resulted in the depletion of LPS available to bind native Factor C, thus causing a reduction in its enzymatic activity (55).

[0184] The percentage competition by rFC of the native Factor C enzyme activity of CAL was calculated based on comparison of enzyme activity with the negative control of pHILD2/151. The following formula was employed: 1$\frac{\begin{array}{c}\left(\mathrm{FU}\mathrm{of}\mathrm{LPS}\mathrm{in}\mathrm{pHILD2}/151+\mathrm{CAL}\right)-\\ \left(\mathrm{FU}\mathrm{of}\mathrm{LPS}\mathrm{in}\mathrm{rFC}+\mathrm{CAL}\right)\times 100\%\end{array}}{\mathrm{FU}\mathrm{of}\mathrm{LPS}\mathrm{in}\mathrm{pHILD2}/151+\mathrm{CAL}}$

[0185] where FU represents fluorescence units.

[0186] The competitive effect of crude rFC on LPS-activated CAL Factor C enzymatic activity was compared with the two partially purified rFC samples based on the above formula. Partial purification of rFC using (NH4)2SO4 precipitation improved its competitive effect from 30% to 60%. Enrichment of rFC through Biomax™-50 improved its inhibitory efficiency to 81%.

[0187] A checkerboard analysis of fixed amount of rFC with variable concentrations of LPS was used to investigate binding efficacy and the ratio of interaction between rFC and LPS molecules. A comparison was made between the binding efficacy of crude rFC and Biomax™-50 rFC to LPS. FIG. 4A shows that increasing amounts of rFC resulted in greater depletion of LPS, leading to an increase in the percentage loss of CAL Factor C activity. On the other hand, regardless of any fixed amount of rFC in the reaction mixture, increasing levels of LPS increased CAL Factor C activity. This indicates that excess LPS was again able to activate CAL Factor C enzyme activity. Even without purification, the crude rFC was able to effectively reduce Factor C activity in CAL by>80%, equivalent to a ratio of 1000:1 molecules of rFC to LPS (FIG. 22A, dashed line).

[0188] With increasing amounts of LPS used over a fixed amount of Biomax™-50 enriched rFC sample, the percentage loss of CAL Factor C activity was effectively maintained at >80% (FIG. 22B). For a >80% reduction in CAL Factor C activity, a ten-fold increase in the binding efficacy was observed between enriched rFC and LPS at a ratio of 100:1 (FIG. 22B, dashed line).

[0189] The Biomax™-50 rFCSN which served as the internal negative control showed little or no effect on the enzymatic activity of CAL Factor C. The rFCEE, having a lower binding affinity of LPS, displayed a markedly lower competitive effect on the enzymatic activity of native Factor C (FIG. 23).

[0190] Only micrograms of the total crude rFC were needed to remove nanograms of LPS, as reflected by ≧80% loss in CAL activity. With partial purification and concentration using Biomax™-50 membrane, the ratio of rFC to LPS for maximal LPS removal improved by ten-fold.

[0191] Interactive binding between rFC and immobilized lipid A using the BIACORE X™ sensor indicates that although background binding was attributed to rFCSN, the overall binding of partially-purified rFC to lipid A gave a net response of approximately 30% of the total immobilized lipid A. Thus, the full-length rFC has an affinity for lipid A such that 30% of lipid A is bound by the partially-purified rFC when the ratio of partially-pure rFC to lipid A is 10:1 on a mass basis. This shows that rFC has affinity for bound lipid A. rFCEE also gave a binding response but again, displayed at a lower affinity of 15% (FIG. 24). For purposes of this application, the terminology “retains lipid A binding activity” indicates an affinity of 10−6 M or lower. Preferably peptides will have a binding affinity of 10−7 M or lower.

[0192] In experiments described in reference 77, rFCES produced in Drosophila cells and purified to homogeneity shows 92% saturation of immobilized lipid A in a BIACORE X™ apparatus under conditions similar to those decribed above, except that the concentration of rFCES was 8 ng/μl. Thus, under these conditions 240 ng of rFCES binds 2.8 μg lipid A giving a ratio of about 1:12 rFCES to lipid A on a mass basis.

EXAMPLE 13

rFC has Bacteriostatic Activity

[0193] The rFC and deletion proteins expressed in P. pastoris in Example 12 were examined for bacteriostatic activity in in vitro cultures.

[0194] The bacteria used for the assay were Escherichia coli ATCC#25922, Salmonella typhimurium ATCC#14028, Pseudomonas aeruginosa ATCC#27853, Klebsiella pneumoniae ATCC#13883 and Staphylococcus aureus ATCC#25923. A colony of each Gram-negative bacterium was inoculated into nutrient broth (Gibco, BRL) and grown at 37° C. until it reached the logarithmic phase of growth. The culture was diluted with nutrient broth to give 1-5×105 cells/ml. Aliquots of 2 ml culture were incubated with 1 mg rFC/ml of culture. Incubation was carried out at 37° C. At time intervals of 0, 2, 4, 6, and 24 h, the bacterial culture was vortexed to break up any agglutinated clumps. After vortexing, each culture was examined under the microscope to ensure homogeneity of bacteria. The culture was serially diluted with 0.85% saline, plated on nutrient agar (Oxoid) and incubated overnight at 37° C. for colony counting.

[0195] Crude full-length rFC and Sephadex™ G-100 enriched rFC truncates rFCEE and rFCSN did not have any inhibitory effect on the growth of the various bacteria used. However, enrichment of full-length rFC by Biomax™-50 ultrafiltration followed by chromatography through Sephadex™ G-100 yielded rFC which inhibited the growth of Gram-negative bacteria such as E. coli, K. pneumoniae, P. aeruginosa, and S. typhimurium (FIG. 25). The enriched rFC sample showed a particularly potent bacteriostatic on K. pneumoniae. This appears consistent with the antibacterial activity found in the cell-free hemolymph of C. rotundicauda (72). Further purification of rFC is expected to improve its bactericidal potential. The bacteriostatic effect was maintained at 100% for 2 h but the effect started to decline at 6 h of incubation, and was completely lost after 24 h. S. aureus, a Gram-positive bacterium, was not inhibited at all by rFC.

[0196] Agglutination of the bacteria was observed within 2 h of incubation with rFC. FIGS. 26A-26E show a typical agglutination reaction exemplified by E. coli This could be attributed to the bacteriostasis of the bacteria because interestingly, there was no agglutination observed with the S. aureus culture which similarly did not show any inhibited growth. Indirectly, the agglutination effect of rFC could be utilized as a rapid detection method and/or for the removal of Gram-negative bacteria from a sample.

[0197] Since LPS is required for reproduction of Gram-negative bacteria (40), it could be envisaged that rFC binds to the lipid A portion of LPS to neutralize its biological activities, causing agglutination which leads to bacteriostasis. This specific binding of rFC to LPS was confirmed by the observation that growth of Gram-positive S. aureus, which does not possess LPS on its outer wall, was not affected by rFC.

EXAMPLE 14

rFC Protects Actinomycin D-Sensitized Mice from LPS Lethality

[0198] Actinomycin D (Sigma) was used to sensitize mice to submicrogram amounts of LPS (65,66). The protective effect of rFC on the mortality of actinomycin D-sensitized/LPS-challenged mice was studied according to protocols previously described (19,67,21). In this in vivo experiment, 500 μl aliquots containing 25 or 50 ng LPS from E. coli 055:B5 (Sigma) and 50 μg rFC produced as in Example 12, or saline, were preincubated at 37° C. for 60 min. The LPS-rFC or LPS-saline mix was combined 1:1 with 250 μg/ml actinomycin D immediately prior to injection. A 0.2 ml volume of this solution containing 25 μg actinomycin D, 5 or 10 ng LPS and 10 μg rFC was injected intraperitoneally into each outbred male Swiss albino (20-25 g) mouse. Groups of 10 mice were used for each replicate set of experiment. The percentage of surviving mice was determined at 72 h.

[0199] An earlier experiment determined that the 50% lethal dose of LPS on Swiss albino mice is 3.16 ng. Amounts of 5 and 10 ng of LPS were therefore used in this in vivo experiment. The protective effect of Sephadex™ G-100 enriched rFC on actinomycin D-sensitized/LPS-challenged mice is shown in FIG. 27. Recombinant Factor C was able to attenuate the toxic effect of LPS, and this resulted in the decreased mortality of the sensitized mice challenged with the rFC-LPS mix. On the other hand, rFCSN that lacks the endotoxin-binding domain did not confer any protection on the LPS-challenged mice. This observation suggests that like LALF (19,67) and human cationic antimicrobial protein CAP18 (21), rFC binds specifically to the biological moiety of LPS to neutralize its lethal effect on mice.

EXAMPLE 15

LPS Binding by rFC Produced in a Baculovirus Host-Vector System

[0200] Recombinant Factor C was produced in Sf9 cells infected with a recombinant baculovirus comprising the CrFC21 cDNA encoding full-length rFC. The recombinant protein was expressed and partially purified by Biomax™ ultrafiltration and gel filtration chromatography over Sephadex™ G-100 as described in reference 76. The recombinant rFC exhibits its normal serine protease activity as shown in that reference. The partially purified, full-length rFC was assayed for activity in binding the LPS obtained from K. pneumoniae, S. minnesota, E. coli and S. typhimurium using the BIACORE X™ system and the experimental conditions indicated in Example 12(8). The data of FIG. 29 show that the rFC binds much more strongly to the LPS from K. pneumoniae then to the LPS from E. coli.

EXAMPLE 16

rFC Produced in Baculovirus-Infected Cells Induces Bacteriostasis and Protects Mice from LPS Lethality

[0201] The rFC produced in the baculovirus-infected Sf9 cells described in Example 15 was assayed for its activity in inducing bacteriostasis and for protective effect in the LPS challenge experiment.

[0202] The bacteriostatic activity of the rFC from Sf9 cells was assessed in the manner described in Example 13. The data in FIG. 30 show that the bacteriostatic effect is observed and that, for E. coli and K. Pneumoniae, it shows similar kinetics to that shown for the rFC obtained from yeast. The bacteriostasis induced for S. typhimurium and P. aeruginosa, on the other hand, was of much shorter duration when rFC produced in the Sf9 cells is used.

[0203] The ability of the rFC produced in recombinant baculovirus-infected Sf9 cells to protect mice from lethal LPS challenge was also tested. The experiment was conducted in the same manner as described in Example 14, except that only 10 μg of the partially purified rFC was administered. In FIG. 31, rFC-Sf9 indicates administration of recombinant rFC together with the indicated amount of LPS; wt-Sf9 indicates that supernatants from Sf9 cells harboring only wild type baculorvirus were used. Consistent with the bacteriostatic effect observed in cultured bacteria, the rFC produced in recombinant baculovirus-infected Sf9 cells was able to protect a significant proportion of the challenged mice from LPS lethality.

[0204] The invention being thus described, various modifications of the materials and methods used in the practice of the invention will be readily apparent to one of ordinary skill in the art. Such modifications are considered to be encompassed by the scope of the invention as described in the claims below.

TABLE 1
A Comparison between binding affinity of Factor C-derived sushi
proteins and other LPS-binding proteins to lipid A.
		Association	Dissociation	Equilibrium
		constant	constant	constant
Proteins	Ligand	(M−1s−1)	(s−1)	(M)	References
Sushi-123	Lipid A E. coli K12	4.01 × 105	1.48 × 10−4	3.691 × 10−10
		5.20 × 105	7.88 × 10−7	1.515 × 10−12
Sushi-1	Lipid A E. coli K12	2.401 × 104	3.64 10−4	1.516 × 10−10
Sushi-3	Lipid A E. coli K12	1.479 × 105	2.031 × 10−4	1.373 × 10−9
Native LALF	Lipid A E. coli K12	3.124 × 104	1.154 × 10−4	3.694 × 10−9
Calionic protein 18 (CAP18)	LPS S. minnesota Re595			5.8 × 10−10	de Haas et al., 1998
Bacterial/Permeability-	Lipid A E. coli J5			4.1 × 10−9	Gazzano-Santoro et al.,
Increasing Protein (BPI)					1992
Recombinant BPI23	Lipid A E. coli J5			2.6-4.3 × 10−9	Gazzano-Santoro et al.,
					1992; 1994
Recombinant BPI21 (rBPI21)	LPS S. minnesota Re595			3.75 × 10−9	de Haas et al., 1998
BPI pep85-99 (15 mer)				1.76 × 10−6	de Haas et al., 1998
Serum amyloid P component	LPS S. minnesota Re595			3.9 × 10−9	de Haas et al., 1998
(SAP)
SAP pep186-200	LPS S. minnesota Re595			1 × 10−5	de Haas et al., 1998
Native LBP	LPS S. minnesota Re595			3.5 × 10−9	Tobias et al., 1995
Recombinant	Lipid A E. coli J5			5.8 × 10−5	Gazzano-Santoro et al.,
LPS-binding protein (LBP)					1994
NH-LBP (aa 1-197)				5.1 × 10−8	Han et al., 1994
Recombinant soluble human	LPS S. minnesota Re595			2.9 × 10−8	Tobias et al., 1995
CD14
Polymixin B				3.3 × 10−7	Vaara M., 1992
Polymixin nonapeptides				1.1-1.3 × 10−6	Vaara and Vijanen, 1985
Limulus endotoxin-binding	LPS E. coli 055:85			6 × 10−6	Minett et al., 1991
protein-protease inhibitor

[0205]

TABLE 2
Comparison of MBC50, MBC90, hemolytic
and cytotoxic activities of Sushi and other
cationic peptides on P. aeruginosa.
			Strains	Killing
Test	MBC50	MBC90	tested	rate*
peptides	(μg/ml)	(μg/ml)	(n)	(min)	Hemolytic activity
S1	≦0.03	≦0.03	30	7	0% at 100 μg/ml
SΔ1	≦0.03	0.06	30	7	0% at 100 μg/ml
S3	≦0.03	≦0.03	30	7	5% at 100 μg/ml
SΔ3	≦0.03	0.25	30	7	35% at 100 μg/ml
SMAP-29	NA	0.38-6.4b	2	NA	67% at 255 μg/mlc
Buforin II	32	64	10	30d	NA
Cecropin I	32	128	10	30d	NA
Indolicin	64	>128	10	60d	NA
Magainin	64	NA	10	60d	NA
II
Nisin	128	NA	10	60d	NA
Ranalexin	>128	NA	10	>60d	NA
a Killing rate for concentration achieving MBC90.
b Only MIC value was reported.
c Hemolysis was done on human erythrocytes. Values were converted from its MW of 3198.
d Time-kill kinetics were tested on Escherichia coli.
NA Data not available
MBC50 Concentration of peptide required to kill 50% of the strains.
MBC90 Concentration of peptide required to kill 90% of the strains.

[0206]

Indicators of LPS-binding, anti-LPS and antimicrobial activities of Factor C & peptides

II

VIII

Stoichi-

Haemo-

ometry

lytic

of

V

Activity

IX

I

Binding

IV

Suppression

at 100

Anti-microbial

Affinity

Coop-

LPS-

of LPS

ug/ml

Potency

for LPS

erativity

Neutral-

Induced

VI

peptide

P. aeruginosa

Factor C

(BIAcore

Hill's

III

isation

cytokine

Mouse

VII

(hu

MBC90 ug/ml (nm)

fragments &

KD

Coef-

CD +

ENC50

TNF-α

Protection

Cytotoxicity

RBC)

or

peptides

(M)

ficient

lip A

(uM)

(50%)

Assay %

(Cell lysis) %

% lysis

MIC∞

SSCrFCES

2.2

0.069

8.5 μM

0.0% at 2 μM

20% lysis at

(76 ug)

109 uM (4 mg/ml)

Sushl 123

3.7 × 10−10

—

—

—

1.5 × 10−12

Sushl 1

1.5 × 10−10

—

—

—

Sushl 3

1.4 × 10−7

—

—

—

S1

1.07 × 10−7

2.42

43% α-H

2.25

—

22% at 75 μg

0

0.03 ug/ml

57% T

(20 nmol)

(8 nM)

S1Δ

6.65 × 10−7

1.08

16% α-H

0.875

53 μM

40% at 75 μg

5-10%

0

0.06 ug/ml

10.8% β

(8.7 nmol)

{close oversize brace}

lysis at

(16 nM)

38.2% T

4 mg/ml

S3

5.85 × 10−7

0.99

90.6% β

1

94 μM

58% at 75 μg

5

≦0.03 ug/m

{close oversize brace}

MBC

9.4% α-H

(10 nmol)

(8 nM)

S3Δ

6.61 × 10−7

0.91

46.5% α-H

0.625

40 μM

100% at

25-30% lysis at

35

0.25 ug/ml

53.5% T

75 μg

4 mg/ml

(63 nm)

(10 nmol)

S4

1.02 × 10−6

0.3

100%

0.01

20 μM

—

3.8 μM

20%

≦0.01 ug/ml

Random

S5

6.31 × 10−6

2.1

100% R

>0.25

>200 μM

—

33 μM

0%

≦0.01 ug/ml

{close oversize brace}

MIC

Random

S6-vg1

1.39 × 10−5

1.0

3.3% α

>0.25

37.5 μM

—

26 μM

5%

≦0.01 ug/ml

Helical

S7-vg2

2.71 × 10−6

1.1

100% R

>0.25

30 μM

—

400 μM

{close oversize brace}

EC50

5%

≦0.01 ug/ml

Random

S8-vg3

2.5 × 10−5

0.9

100% α

0.03

20 μM

—

22 μM

100%

≦0.01 ug/ml

{close oversize brace}

MBC

Helical

S9-vg4

1.73 × 10−5

1.0

100% α

>0.25

25 μM

—

35 μM

75%

≦0.01 ug/ml

Helical

References

[0207] Articles of the patent and scientific periodical literature have been cited throughout this document. Each such article is hereby incorporated in its entirety by reference by such citation.

[0208] 1. Karima R., Matsumoto, S., Higashi, H. and Matsushima, K. (1999) The molecular pathogenesis of endotoxic shock and organ failure. Mol. Med. Today 5(3) 123-132.

[0209] 2. Kreutz, M., Ackermann, U., Hauschildt, S., Krause, S. W., Riedel, D., Bessler, W. and Andreesen, R. (1997) A comparative analysis of cytokine production and tolerance induction by bacterial lipopeptides, lipopolysaccharides and Staphyloccous aureus in human monocytes. Immunology 92(3), 396-401.

[0210] 3. Tracey, K. J. and Lowry, S. F. (1990) The role of cytokine mediators in septic shock. Adv. Surg. 23, 21-56.

[0211] 4. Zahringer, U., Lindner, B. and Rietschel, E. T. (1994) Molecular structure of lipid A, the endotoxic center of bacterial lipopolysaccharides. Adv. Carbohydr. Chem. Biochem. 50, 211-76.

[0212] 5. Armstrong, P. B. and Rickles, F. R. (1982) Endotoxin-induced degranulation of the Limulus amebocyte. Exp. Cell Res. 140, 15-24.

[0213] 6. Levin, J., Tomasulo, P. A. and Oser, R. S. (1970) Detection of endotoxin in human blood and demonstration of an inhibitor. J. Lab. Clin. Med. 75, 903-911.

[0214] 7. Novitsky, T. J. (1994) Limulus amoebocyte lysate (LAL) detection of endotoxin in human blood. J. Endotoxin Res. 1, 253-263.

[0215] 8. Muta, T. and Iwanaga, S. (1996) Clotting and immune defense in Limulidae. Prog. Mol. Subcell. Biol. 15, 154-189.

[0216] 9. Nakao, A., Tasul, M., Shen, S. and Takagi, H. (1995) Heparin effects on superoxide production by neutrophils. Eur. Surg. Res. 27(4), 216-221.

[0217] 10. Ding, J. L., Navas, M. A. A. and Ho, B. (1925) Molecular cloning and sequence analysis of Factor C cDNA from the Singapore horseshoe crab, Carcinoscorpius rotundicauda. Mol. Marine Biol. Biotech. 4(1), 90-103.

[0218] 11. Roopashree, S. D., Chai, C., Ho, B. and Ding, J. L. (1995) Expression of Carcinoscorpius rotundicauda Factor C cDNA. Biochem. Mol. Biol. Intl. 35(4), 841-849.

[0219] 12. Roopashree, S. D., Ho, B. and Ding, J. L. (1996) Expression of Carcinoscorpius rotundicauda Factor C in Pichia pastoris. Mol. Marine Biol. Biotech. 5(3), 334-343.

[0220] 13. Pui, A. W. M., Ho, B. and Ding, J. L. (1997) Yeast recombinant Factor C from horseshoe crab binds endotoxin and causes bacterostasis. J. Endotoxin Res. 4(6), 391-400.

[0221] 14. Clare, J. J. et al. And Henwood, C. A. (1991) Production of mouse epidermal growth factor in yeast: high level secretion using Pichia pastoris strains containing multiple gene copies. Gene 105, 205-212.

[0222] 15. Hussein, A. S., Chacon, M. R., Smith, A. M., Tosado-Acevedo, R., Selkirk, M. E. (1999) Cloning, expression, and properties of a nonneuronal secreted acetylcholinesterase from the parasitic nematode Nippostrongylus brasiliensis. J Biol. Chem. 274(14), 9312-9319.

[0223] 16. Muta, T., Miyata, T., Misumi, Y., Tokunaga, F., Nakamura, T., Toh, Y., Ikehara, Y. and Iwanaga, S. (1991) Limulus factor C. An endotoxin-sensitive serine protease zymogen with a mosaic structure of complement-like, epidermal growth factor-like, and lectin-like domains. J. Biol. Chem. 266, 6554-6561.

[0224] 17. Righetti, P. G. and Giaffreda, E. (1994) Immobilized buffers for isoelectric focusing: from gradient gels to membranes. Electrophoresis 15(8-9), 1040-1043.

[0225] 18. Marra, M. N., Wilde, C. G., Griffith, I. E., Snable, J. L. and Scott R. W. (1990) Bactericidal/permeability-increasing protein has endotoxin-neutralizing activity. J. Immunol. 144, 662-666.

[0226] 19. Warren H. S. Glennon, M. L., Wainwright, N., Amato, S. F., Black, K. M., Kirsch, S. J., Riveau, G. R., Whyte, R. I., Zapol, W. M. and Novitsky, T. J. (1992) Binding and neutralization of endotoxin by Limulus Antilipopolysaccharide Factor. Infect. Immun. 60(6), 2506-2513.

[0227] 20. Appelemelk, B. J., An, Y. Q., Geerts, M. Thijs, B. G., DeBoer, H. A., MacLaren, D. M., DeGraaff, J. and Nuijens, J. H. (1994) Lactoferrin is a lipid A-binding protein. Infect. Immun. 62, 2628-2632.

[0228] 21. Larrick, J. W., Hirata, M., Balint, R. F., Lee, J., Zhong, J. and Wright, S. C. (1995) Human CAP18: a novel antimicrobial lipopolysaccharide-binding protein. Infect. Immun. 63(4), 1291-1297.

[0229] 22. O'Riordan, J. F., Goldstick, T. K., Ditzel, J. and Ernest, J. T. (1983) Characterization of oxygen-hemoglobin equilibrium curves using nonlinear regression of the Hill equation: parameter values for normal human adults. Adv Exp Med Biol 159, 435-444.

[0230] 23. Hill, A. V. (1910) The combinations of haemoglobin with oxygen and with carbon monoxide. I. J. Physiol. (Lond.) 40, 4-8.

[0231] 24. Zhang, G. H., Mann, D. M. and Tsai, C. M. (1999) Neutralization of endotoxin in vitro and in vivo by a human lactoferrin-derived peptide. Infect. Immun. 67(3), 1353-1358.

[0232] 25. Tsuchiya, S., Kobayashi, Y., Goto, Y., Okumura, H., Nakae, S., Konno, T. and Tada, K. (1982) Induction of maturation in cultured human monocytic leukemia cells by a phorbol diester. Cancer Res. 42(4), 1530-1536.

[0233] 26. Ried, C., Wahl, C., Miethke, T., Wellnhofer, G., Landgraft, C., Schneider-Mergener, J. and Hoess, A. (1996) High affinity endotoxin-binding and neutralizing peptides based on the crystal structure of recombinant Limulus Antilipopolysaccharide Factor. J. Biol. Chem. 45, 28120-28127.

[0234] 27. Cory, A. H., Owen, T. C., Barltrop, J. A. and Cory, J. G. (1991) Use of an aqueous soluble tetrazolium/formazan assay for cell growth assays in culture. Cancer Commun. 3, 207-212.

[0235] 28. Riss, T. L. and Moravec, R. A. (1992) Comparison of MTT, XTT, and a novel tetrazolium compound MTS for in vitro proliferation and chemosensitivity assays. Mol. Biol. Cell 3 (Suppl.), 184a.

[0236] 29. Lehmann, V., Freudenberg, M. A. and Galanos, C. (1987) Lethal toxicity of lipopolysaccharide and tumor necrosis factor in normal and D-galactosamine-treated mice. J. Exp. Med. 165, 657-663.

[0237] 30. Laub, P. B. and Gallo, J. M. (1996) NCOMP—a windows-based computer program for noncompartmental analysis of pharmacokinetic data. J Pharm. Sci. 85(4), 393-5.

[0238] 31. Galanos, C., Freudenberg, M. A. and Reutter, W. (1979) Galactosamine-induced sensitization to the lethal effects of endotoxin. Proc. Natl. Acad. Sci. USA 76, 5939-5943.

[0239] 32. Kaplan, E. L. and Meier, P. (1958) Nonparametric estimation from imcomplete observations. J. Am. Stat. Assoc. 53, 457-481.

[0240] 33. Andrea, G., Cirioni, O., Barchiesi, F, Prete, M S D and Scalise G. (1999) Antimicrobial activity of polycationic peptides. Peptides. 20: 1265-1273.

[0241] 34. Agerberth, B., Gunne, H., Odeberg, I., Kogner, P., Boman, H G and Gudmundsson, G (1995). FALL-39, a putative human peptide antibiotic is cysteine-free and expressed in bone marrow and testis. Proc. Natl. Acad. Sci. USA. 92: 195-199.

[0242] 35. Barbara, S., Benincasa, M., Risso, A, Zanetti, M and Gennaro, R. (1999). SMAP-29: a potent antibacterial and antifungal peptide from sheep leukocytes. FEBS Letts. 463: 58-62.

[0243] 36. Teshima, T., Ueky, Y., Nakai, T. and Shiba, T. (1986). Structure determination of lepidopteran, self-defense substance produced by silkworm. Tetrahedron. 42: 829-834.

[0244] 37. Sawa, T., Kurahashi, K, Ohara, M., Gropper, M A., Doshi, V., Larrick, J W. And Wiener-Kronish, J P. (1998). Evaluation of antimicrobial and lipopolysaccharide-netralising effects of a synthetic CAP-18 fragment against Pseudomonas aeruginosa in a mouse model. Antimicrob. Agents Chemother. 42: 3269-3275.

[0245] 38. Bone, R C. (1991). The pathogenesis of sepsis. Ann. Intern. Med. 115: 457-460.

[0246] 39. Parrillo, J E., Parker, M M, Nathanson, C., Cunmnion, A F., Ognibene F P. (1990). Septic shock in humans. Advances in the understanding of pathjogenesis, cardiovascular, dysfunction and therapy. Ann. Intern. Med. 113: 227-237.

[0247] 40. Rietschel E. T. et al., Sci Am Aug. 1992; 26-33.

[0248] 41. Tobias P. S. et al., J Exp Med 1986; 164: 777-793.

[0249] 42. Marra M. N. et al., J Immunol 1992; 148: 532-537.

[0250] 43. Tobias P. S. et al., J Biol Chem 1989; 264: 10867-10871.

[0251] 44. Elsbach P. et al., Curr Opin Immunol 1993; 5: 103-107.

[0252] 45. Rogy M. A. et al., J Clin Immun 1994; 14: 120-133.

[0253] 46. Tanaka S. et al., Biochem Biophys Res Commun 1982; 105: 717-723.

[0254] 47. Morita T. S. et al., J Biochem 1985; 97: 1611-1620.

[0255] 48. Aketagawa J. et al., J Biol Chem 1986; 261: 7357-7365.

[0256] 49. Muta T. et al., J Biochem 1987; 101: 1321-1330.

[0257] 50. Alpert G. et al., J Infect Dis 1992; 165: 494-500.

[0258] 51. Fletcher M. A. et al., J Surg Res 1993; 55: 147-154.

[0259] 52. Garcia C. et al., Crit Care Med 1994; 22: 1211-1218.

[0260] 53. Nakamura T. et al., J Biol Chem 1988; 263: 16709-16713.

[0261] 54. Miyata T. et al., J Biochem 1989; 106: 663-668.

[0262] 55. Saito T. et al., J Biochem 1995; 117: 1131-1137.

[0263] 56. Levin J. et al., Bull John Hopkins Hosp 1964; 115: 265-274.

[0264] 57. Nakamura T. et al., Eur J Biochem 1986; 154: 511-521.

[0265] 58. Tokunaga F. et al., Eur J Biochem 1987; 167: 405-416.

[0266] 59. Ho B., Microbios Lett 1983; 24: 81-84.

[0267] 60. Roopashree S. D. et al., Biochem Mol Biol Intl 1995; 35: 841-849.

[0268] 61. Ding J. L. et al., J Endotoxin Res 1997; 4: 33-43.

[0269] 62. Bradford M. M. et al., Anal Biochem 1976; 72: 248-254.

[0270] 63. Laemmli U. K. et al., Nature 1970; 227: 680-684.

[0271] 64. Iwanaga S. et al., Japan Patent Agency Official Bulletin 1980; 557-108018.

[0272] 65. Pieroni R. E. et al., Proc Soc Exp Biol Med 1971; 133: 790-794.

[0273] 66. Brown D. E. et al., J Infect Dis 1982; 146: 746-750.

[0274] 67. Kloczewiak M. et al., J Infect Dis 1994; 170; 1490-1497.

[0275] 68. Nakamura T. et al., J Biochem 1988; 103: 370-374.

[0276] 69. Sandri M. et al., Anal Biochem 1993; 213: 34-39.

[0277] 70. Tsai C. M. et al., Anal Biochem 1982; 119: 115-119.

[0278] 71. Miura Y. et al., J Biochem 1992; 112: 476-481.

[0279] 72. Yeo D. S. A. et al., Microbios 1993; 73: 45-58.

[0280] 73. U.S. Pat. No. 5,716,834

[0281] 74. U.S. Pat. No. 5,712,144

[0282] 75. U.S. patent application Ser. No. 08/877,620

[0283] 76. U.S. patent application Ser. No. 09/081,767

[0284] 77. U.S. Patent Application No. 60/106,426

[0285] 78. U.S. patent application Ser. No. 09/201,786

[0286] 79. “Remington, the Science and Practice of Pharmacy”, 19th Edition, c. 1995 by the Philadelphia College of Pharmacy and Science.

Claims

1. A method for treating gram negative bacterial infection in a subject comprising administering an amount of recombinant Factor C effective for producing bacteriostasis.

2. The method of claim 1, wherein said recombinant Factor C is produced by a yeast host cell or by an insect host cell.

3. The method of claim 2, wherein said recombinant Factor C lacks serine protease activity but retains lipid A binding activity.

4. The method of claim 1, wherein said recombinant Factor C is encoded by a nucleic acid that hybridizes to a nucleic acid having the sequence of SEQ ID NO:1 or SEQ ID NO:3 under stringent conditions.

5. The method of claim 1, wherein the recombinant Factor C has the amino acid sequence of SEQ ID NO:2 or of SEQ ID NO:4, has residues 1-766 of SEQ ID NO:4, residues 29-330 of SEQ ID NO:4, residues 29-201 of SEQ ID NO:4, or residues 264-330 of SEQ ID NO:4, or has three sushi domains linked by random amino acid sequences.

6. The method of claim 5, wherein the recombinant Factor C has at least amino acids 60-70, 170-185, and 270-280 of SEQ ID NO:4.

7. The method of claim 1, wherein the Gram negative bacterial infection comprises bacteria selected from the group consisting of K. pneumoniae and E. coli.

8. The method of claim 6, wherein the subject is a mammal.

9. The method of claim 1, wherein the amount administered is within the range of 0.1 to 1.0 mg per kg body weight of the subject.

10. An isolated polypeptide comprising a lipopolysaccharide binding domain of a Factor C protein.

11. The polypeptide of claim 10, wherein said polypeptide is substantially free of hemolytic activity but retains lipid A binding activity.

12. The polypeptide of claim 10, wherein the lipopolysaccharide binding domain of Factor C protein is selected from the group consisting of: amino acids 1-333 of a Factor C protein; at least one member selected from the group consisting of a sushi 1 domain of a Factor C protein, a sushi 2 domain of a Factor C protein, and a sushi 3 domain of a Factor C protein; a sushi-1 peptide; a sushi-1Δ peptide; a sushi-3 peptide; and a sushi-3Δ peptide.

13. The polypeptide of claim 10, further comprising a secretory signal sequence of a vitellogenin protein.

14. The polypeptide of claim 13 that is purified SSCrFCES.

15. The polypeptide of claim 10, comprising a member selected from the group consisting of a sushi-1 peptide, a sushi-1Δ peptide, a sushi-3 peptide, and a sushi-3Δ peptide.

16. The polypeptide of claim 10, further comprising a reporter protein or an affinity tag.

17. The polypeptide of claim 16, comprising a reporter protein selected from the group consisting of green fluorescent protein (GFP), alkaline phosphatase, a peroxidase, and a luciferase.

18. The polypeptide of claim 17, comprising a member selected from the group consisting of SSCrFC-sushi-1-GFP, SSCrFC-sushi-3-GFP, and SSCrFC-sushi-1,2,3-GFP.

19. The polypeptide of claim 16, comprising an affinity tag selected from the group consisting of polyhistidine or biotin.

20. A method for treating sepsis caused by a gram negative bacterial infection comprising administering a polypeptide of claim 10 to a subject in an amount effective to bind lipopolysaccharide of said gram negative bacteria and ameliorate inflammatory response to said lipopolysaccharide.

21. The method of claim 20, wherein said polypeptide is a member selected from the group consisting of a sushi-1 peptide, a sushi-1Δ peptide, a sushi-3 peptide, and a sushi-3Δ peptide.

22. The method of claim 21, wherein said polypeptide is substantially free of hemolytic activity but retains lipid A binding activity.

23. The method of claim 21, wherein the gram negative bacterial infection comprises bacteria select from the group consisting of P. aeruginosa, K. pneumoniae, and H. pylori.

24. The method of claim 21, wherein the subject is a mammal.

25. A pharmaceutical composition comprising a therapeutically effective amount of a polypeptide of claim 10 and a pharmaceutically acceptable carrier for topical formulation.

26. A method for treating or preventing infection of a wound by gram negative bacteria comprising administering the composition of claim 25 to said wound.

27. A method for the detection of gram negative bacteria or of lipopolysaccharide in a sample comprising contacting a sample to be assayed for the presence of said gram negative bacteria or lipopolysaccharide with the polypeptide of claim 12 wherein presence of gram negative bacteria or lipopolysaccharide is indicated by a complex between said gram negative bacteria or lipopolysaccharide and said polypeptide of claim 12.

28. The method of claim 27, comprising an in situ histologic assay.

29. The method of claim 27, which is a solution assay.

30. The method of claim 27, wherein said polypeptide is immobilized.

31. The method of claim 27, wherein gram negative bacteria or lipopolysaccharide of said sample is immobilized.

32. A method for the detection of whole or fragmentary gram negative bacteria or of lipopolysaccharide in a sample comprising contacting a sample to be assayed for the presence thereof with a polypeptide of claim 10, further comprising a reporter protein.

33. The method of claim 32, wherein said reporter protein comprises a green fluorescent protein.

34. The method of claim 33, wherein said sample comprises tissues or cells and said polypeptide comprises a member selected from the group consisting of SSCrFC-sushi-1-GFP, SSCrFC-sushi-3-GFP, and SSCrFC-sushi-1,2,3-GFP.

35. A method for preserving a sample from contamination by gram negative bacteria comprising adding a polypeptide of claim 10 to said sample in an amount effective for preventing the growth of said gram negative bacteria.

36. A method for purifying a sample by removal of endotoxin comprising immobilizing a polypeptide of claim 12 on an insoluble substrate, contacting said sample with the immobilized polypeptide, and separating said sample from said immobilized polypeptide.

37. An isolated nucleic acid encoding the polypeptide of claim 10.

38. An isolated nucleic acid comprising a nucleic acid encoding a lipopolysaccharide binding portion of a Factor C protein of a horseshoe crab selected from the group consisting of; amino acids 1-333 of a Factor C protein; at least one member selected from the group consisting of a sushi 1 domain of a Factor C protein, a sushi 2 domain of a Factor C protein, and a sushi 3 domain of a Factor C protein; a sushi-1 peptide; a sushi-1Δ peptide; a sushi-3 peptide; and a sushi-3Δ peptide.

39. The isolated nucleic acid of claim 38, further comprising a nucleic acid encoding a secretion signal sequence of a vitellogenin protein.

40. The isolated nucleic acid of claim 38, further comprising a nucleic acid that encodes a reporter protein or an affinity tag fused to the nucleic acid encoding the lipopolysaccharide binding portion of a Factor C protein.

41. The isolated nucleic acid of claim 39, comprising a nucleic acid encoding a reporter protein selected from the group consisting of green fluorescent protein, alkaline phosphatase, a peroxidase, and a luciferase.

42. The isolated nucleic acid of claim 39, comprising a nucleic acid encoding an affinity tag selected from the group consisting of a polyhistidine sequence or biotin.

43. The isolated nucleic acid of claim 42, wherein the nucleic acid encodes a member selected from the group consisting of SSCrFCES, SSCrFC-sushi-1-GFP, SSCrFC-sushi-3-GFP, and SSCrFC-sushi-1,2,3-GFP.

44. A method for producing an isolated lipopolysaccharide binding protein comprising: i) culturing a host cell transformed with the isolated nucleic acid of claim 38 to produce said lipopolysaccharide binding protein in a culture medium; and ii) isolating said lipopolysaccharide binding protein from said culture medium.

45. The method of claim 44, wherein said isolating step ii) comprises ultrafiltering the culture medium with a 10 kDa cutoff ultrafiltration membrane and preparative isoelectric focussing.

45. A recombinant polypeptide produced by the process of claim 44.