CA2323630A1 - Use of bactericidal/permeability increasing protein or biologically active analogs thereof to treat lipopolysaccharide associated gram negative infections - Google Patents

Abstract

The present invention provides a method of inhibiting lipopolysaccharide (LPS)-mediated stimulation of cells. This method comprises contacting the cells, in the presence of a cell-stimulating amount of lipopolysaccharide, with Bactericidal/Permeability Increasing Protein (BPI) in an amount effective to inhibit cell stimulation.

Description

USE OF BACTERICIDAL/PE ARIT.rrv INCREASING PROTEIN OR

LIPOPOLYSACCIiAR_1DE ASSOCLATED GReM NE ATIVE INFEG'TIONS
Background of the Invention o Gram negative infections are a major cause of morbidity and mortality especially in hospitalized and immunocompromised patients. [Duma, RJ., Am. J. of Med., 78 (Suppl. 6A): 154-164 (1985); and Kreger B.E., D.E. Craven and W.R. McCabe, Am. J. Med., 68: 344-355 ( 1980)]
~ 5 Although available antibiotics are effective in containing the infection, they do nothing to neuualize the pathophysiological effects associated with lipopolysaccharide (LPS). LPS, or endotoxin, is a major component of the outer membrane of gram negative bacteria and is released when the organisms are 20 Ivsed. [Ahenep, J.L.. and KA. Morgan, J. Infect. Dis., 150 (3): 380-388 ( 1984)]
L.PS released during antibiotic therapy is a potent stimulator of the inflammatory response. Many detrimental effects of L.PS in vivo result from soluble mediators 25 released by inflammatory cells. [Morrison D.C. and RJ. Ulevich, Am. J.
Pathol., 93 (2): 527-617 (1978)] L.PS induces the release of mediators by host inflammatory cells which may ultimately result in disseminated intravascular coagulation (DIC), adult respiratory disuess syndrome CARDS), renal failure, and irreversible shock.

~, WO 90/09183 ~ PCI'/US90/00837 Monocytes and neutrophilic granulocytes play a key role in host defense against bacterial infections and also participate in the pathology of endotoxemia.
These cells ingest and kill microorganisms intracellularly and also respond to LPS
in vivo and in vitro by releasing soluble proteins with microbicidal, proteolytic, opsonic, pyrogenic, complement activating and tissue damaging effects. Tumor necrosis factor (TNF), a cytokine released by LPS stimulated monocytes misics some of the toxic effects of LPS in vivo. Injecting animals with TNF causes fever, shock and alterations in glucose metabolism. TNF is also a potent stimulator of neutrophils.
Soluble LPS causes decreased neutrophil chemotaxis, increased adhesiveness, elevated hexose monophosphate shunt activity and O= radical. production, upregulation of surface receptors for complement, and release of granule proteins into the surrounding medium. [Morrison and Ulevich ( 1978)]
Both specific and azurophil compartments degranulate in response to LPS.
[Bannatyne, R.M., N.M. Harnett, KY. Lee and W.D. Rigger, J. Infect. Dis., 156 (4): 469-474 ( 1977)] Azurophil proteins released in response to LPS may be both 2o harmful and beneficial to the host. Neutrophil elastase causes degradation of protease inhibitors responsible for suppressing the coagulation cascade. This results in coagulopathies such as disseminated intravascular coagulation, a potentially lethal consequence of endotoxemia. Azurophil granules also contain 25 bactericidal molecules such as myeloperoxidase and Bactericidal/Permeability Increasing Factor (BPI).
Rabbit BPI was first discovered in 1975. [Weirs, J., R.C. Franson, S.
Becherdite, K Schmeidler, and P. Elsbach, J. Clin. Invest., 55:33 (1975)] BPI was isolated from human neutrophils in 1978. [Weirs, J., P. Elsbach, I. Olson and H. Odeberg, J.
Biol. Chem, 253 (8): 2664-2672 (1978)].

.. WO 90/09183 PCT/US90/00837 In 1984 a 57 kD protein with similar properties was isolated from human neutrophils. [Shafer, W.M., C.E. Martin and J.K Spitznagel, Infect. Immun., 45:29 ( 1984)] This protein is identical to BPI by N-Terminal sequence amino acid composition, molecular weight and source. Although, the authors were unable to reproduce the chromatographic isolation procedure used by Elsbach, et al. and Weiss, et al.
Human BPI is a 57 kD protein which binds to the outer membrane of susceptible gr~ negative bacteria. [Weiss, et al. ( 1978)] The fact that BPI is a Lipid A
binding protein is evidenced by: (1) rough strains of bacteria are more sensitive to both bactericidal and permeability increasing activities of BPI [Weirs, J., M.
Hutzler and L Kao, Infect. Immun., 51:594 ( I986)]; (2) mutations in Lipid A
cause decreased binding and increase resistance to bactericidal activity of both ~ 5 polymyxin B and BPI [Farley, M.M., W.M. Shafer and J.K Spitznagel, Infect.
Immun., 56:1589 ( 1988)]; (3) BPI competes with polymyxin B for binding to ~
~himurium [Farley 1988]; (4) BPI has sequence homology and immunocrossreactivity to another LPS binding protein Lipopolysaccharide Binding Protein (LBP). LBP-LPS complexes have been shown to stimulate the oxidative burst on neutrophils in response to formylated peptides. High density lipoprotein (HDL), another LPS binding protein, found in human serum in complex with LPS
does not show the stimulatory effect on neutrophils. BPI binding disrupts LPS
s~~e, alters microbial permeability to small hydrophobic molecules and causes cell death (Weirs, et al., 1978). BPI kills bacteria under physiologic conditions of pH and ionic strength in vitro indicating that it may be active in vivo outside the low pH environment of the phagolysosome. All of the bactericidal and permeability increasing activities of BPI are present in the N-terminal 25kD
fragment of the protein. [Ooi, C.E., J. Weirs, P. Elsbach, B. Frangione, and B.
Marrion, J. Biol. Chem., 262: 14891 ( 1987)] Prior to the subject invention.

yWp 90/09183 , PCT/US90/00837 however, it has been understood that the beneficial effects of BPI are limited to its bactericidal effects.
Despite improvements in antibiotic therapy, morbidity and mortality associated with endotoxemia remains high. Antibiotics alone are not effective in neutralizing the toxic effects of LPS. Therefore, the need arises for an adjunct therapy with direct LPS neutralizing activity. Current methods for treatment of endotoxemia use antibiotics and supportive care. Most available adjunct therapies treat symptoms of endotoxic shock such as low blood pressure and fever but do not inactivate endotoxin. Other therapies inhibit inflammatory host responses to LPS.
As indicated below, present therapies have major limitations due to toxicity, immunogenicity, or irreproducible efficacy between animal models and human trials.
~ 5 Polymyxin B is a basic polypeptide antibiotic which has been shown to bind to, and structurally disrupt, the most toxic and biologically active component of endotoxin, Lipid A. Polymyxin B has been shown to inhibit LPS activation of neutrophil granule release in vitro and is an effective treatment for gram negative sepsis in humans. However, because of its systemic toxicity, this drug has limited use except as a topical agent.
Combination therapy using antibiotics and high doses of methylprednisolone sodium succinate (MPSS) has been shown to prevent death in an experimental model of gram negative sepsis using dogs. Another study using MPSS with antibiotic in a multicenter, double blind, placebo-controlled, clinical study in 2'_'3 patients with clinical signs of systemic sepsis concluded that mortality was not significantly different between the treatment and placebo groups. Further, the investigators found that resolution of secondary infection within 14 days was significantly higher in the placebo group.

WO 90/09183 PC?/US90/00837 A relatively new approach to treatment of endotoxemia is passive immunization with endotoxin neutralizing antibodies. Hyperimmune human immunoglobulin against E.E. Coli JS has been shown to reduce mortality in patients with gram negative bacteremia and shock by SO%. Other groups have shown promising 5 results in animal models using mouse, chimeric, and human monoclonal antibodies. Although monoclonal antibodies have advantages over hyperimmune sera, e.g. more consistent drug potency and decreased transmission of human pathogens, there are still many problems associated with administering l~unoglobulin to neutralize LPS. Host responses to the immunoglobulins themselves can result in hypersensitivity. Tissue damage following complement activation and deposition of immune complexes is another concern in the use of therapies involving anti-endotoxin antibodies in septic patients. Also, immunoglobulins are large molecules, especially the pentameric IgMs currently in ~ 5 clinical trials, and are rapidly cleared by the reticuloendothelial system, diminishing the half-life of the drug.
Endotoxins elicit responses which are beneficial as well as damaging to the host.
Endotoxemia induces production of LPS binding proteins from the liver and causes release of microbicidal proteins from leukocytes. In applicants' studies of neutrophil proteins involved in host defense, it has been determined that one of these proteins, BPI, is not only a potent microbiridal agent in vitro, but it also interferes with the ability of LPS to stimulate neutrophils. Specifically, it has been demonstrated that BPI binds to soluble LPS and neutralizes its ability to activate neutrophils. Accordingly, this invention provides a therapeutic method for the treatment of LPS toxicity in gram negative-septicemia.

> WO 90/09183 PC1"/US90/00837 Summary of the Invention The present invention provides a method of inhibiting lipopolysaccharide (LPS)-mediated stimulation of cells. This method comprises contacting the cells, in the presence of a cell-stimulating amount of lipopolysaccharide, with Bactericidal/Permeability Increasing Protein (BPI) in an amount effective to inhibit cell stimulation.
0 The invention further provides a method of treating a gram negative bacterial infection. This method comprises contacting the bacterial infection with purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit LPS-mediated stimulation of cells and thereby treat the bacterial infection.
Additionally, the present invention provides for a composition for treatment of a gram negative bacterial infection. This composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit LPS-mediated stimulation of cells and a suitable carrier.
The present invention additionally provides a method of treating a subject suffering from endotoxin-related shock caused by a gram negative bacterial infection which comprises administering to the subject an amount of BPI
effective to combat the gram negative infection and treat the subject so as to alleviate the endotoxin-related shock.
Further, the invention provides a method of treating a subject suffering from disorder involving disseminated intravascular coagulation. The method comprises administering to the subject an amount of BPI effective to alleviate the symptoms of disseminated intravascular coagulation and thereby treat the subject.

a WO 90/09183 PCT/US90/00837 Fun per, the present invention provides a method of treating a subject suffering from endotoxemia caused by a gram negative infection which comprises administering to the subject an amount of BPI effective to combat the gram negative bacterial infection and treat the subject suffering from endotoxemia.
As used herein endotoxemia means a condition in which the blood contains poisonous products, either those produced by the body cells or those resulting from microorganisms, i.e. gram negative bacteria.
The invention also provides a method of treating a subject suffering from endotoxin-related anemia caused by a gram negative bacterial infection which comprises administering to the subject an amount of BPI effective to combat the gram negative bacterial infection and treat the subject so as to alleviate endotoxin-related anemia.
The present invention further provides a method of treating a subject suffering from endotoxin-related leukopenia caused by a gram negative bacterial infection.
The method comprises administering to the subject an amount of BPI effective to combat the gram negative bacterial infection and treat the subject so as to alleviate endotoxin-related leukopenia. Further, the invention includes a method of treating a subject suffering from endotoxin-related thrombocytopenia caused by a gram negative bacterial infection which comprises administering to the subject ~~ ~°~t of BPI effective to combat the gram negative bacterial infection and treat the subject so as to alleviate endotoxin-related thrombocytopenia.
The invention also provides a method of inhibiting a pyrogen which comprises contacting the pyrogen with an amount of BPI so as to inhibit the pyrogen.

,WO 90/09183 ~ PCT/US90/00837 Furthermore, the invention also includes a method of inhibiting lipopolysaccharide-mediated tumor necrosis factor production by cells which comprises contacting the cells in the presence of a cell-stimulating amount of lipopolysaccharide, with BPI in an amount effective to inhibit lipopolysaccharide-mediated tumor necrosis factor production by cells.
The invention also includes a method of inhibiting gram negative bacteria-mediated tumor necrosis factor production by cells which comprises contacting the gram negative bacteria with BPI in an amount effective to inhibit gram negative-mediated tumor necrosis factor production by cells.
Moreover, the invention includes a composition for the treatment of a subject suffering from endotoxin-related shock. The composition comprises a purified BPI
or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related shock and a suitable carrier.
Further, the invention includes a composition for the treatment of a subject suffering from disseminated intravascular coagulation. The composition comprises 2~ a purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from disseminated intravascular coagulation and a suitable carrier.
~~ mention also includes a composition for the treatment of a subject suffering from endotoxemia comprising a purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxemia and a suitable carrier.
.
The invention additionally provides a composition for the treatment of a subject suffering from endotoxin-related anemia comprising purified BPI or a biologicaliv activ~ polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related anemia and a suitable carrier.
Additionally, the invention provides a composition for the treatment of a subject suffering from endotoxin-related leukopenia. The composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related leukopenia and a suitable carrier.
Further provided is a composition for the treatment of a subject suffering from endotoxin-related thrombocytopenia. The composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related thrombocytopenia and a suitable carrier.
As used herein endotoxin-related leukopenia is a condition associated with a gram negative bacterial infection, the manifestation of which is a decrease below the normal number of leukocytes in the peripheral blood. Moreover, as used herein endotoxin-related thromborytopenia is a condition associated with a gram negative bacterial infection, the manifestation of which is a decrease below the normal number of thrombocytes.
Also, the invention provides a composition for inhibiting lipopolysaccharide-mediated tumor necrosis factor production by cells comprising purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit lipopolysaccharide-mediated tumor necrosis factor production by cells and a suitable carrier. Moreover, the invention provides a composition for inhibiting gram negative bacteria-mediated tumor necrosis factor production by cells comprising purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit gram negative bacteria-mediated tumor necrosis factor production by cells and a suitable carrier.

'WO 90/09183 , PCT/US90/00837 The invention provides a composition for inhibiting a pyrogen. Il~e composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit a pyrogen.
Additionally, the invention provides a method of preventing a symptom associated 5 with a gram negative bacterial infection in a subject which comprises administering to the subject an amount of Bactericidal/Permeability Increasing Protein effective to prevent the gram negative bacterial infection and thereby prevent the symptom.

10 Moreover, also provided is a method of preventing a disorder involving disseminated intravascular coagulation in a subject which comprises administering to the subject an amount of Bactericidal/Permeability Increasing Protein effective to prevent the symptoms of disseminated intravascular coagulation and thereby preventing the disorder.
Finally, the invention provides a method of isolating and recovering purified Bartericidal/Permeability Increasing Protein which comprises: (a) obtaining a crude sample of Bactericidal/Permeability Increasing Protein; and (b) separating 2~ the crude sample by column chromatography using de-pyrogenated solutions thereby isolating and recovering purified Bactericidal/Permeability Increasing Protein.

Figure la:Mean fluorescence intensity of CR1 on freshly isolated neutrophils was measured by FACS analysis. Cells were stimulated with varying doses of E. Coli 0111:B4 LPS as described in Materials and Methods. Since mean fluorescence intensity varies between individuals, the data is expressed as percent of the maximum response observed. Data shown represents the mean + /- Standard Error of three experiments.
Figure lb:O111:B4 LPS (10 ng/ml) was preincubated with varying doses of crude azurophil extract for 30 minutes at 37 ~ C prior to testing for neutrophil stimulation. Data shown represents the mean + /- Standard Error of duplicates from a representative experiment. Values are expressed as % inhibition of the response to LPS alone.
Figure 2:Crude azurophil extract was separated by reverse phase HPLC. Each peak was collected manually and protein concentrations were determined by amino acid analysis. An aliquot ( l,ug) of each peak was dried in the presence of low endotoxin BSA, then redried in the presence of pyrogen free 0.1% acetic acid.
Data shown represent the mean + /- Standard Error of duplicates from a representative experiment.
Figure 3a:BPI purified by size exclusion followed by ration exchange HPLC was subjected to reverse phase HPLC and fractions were tested for LPS inhibitory activity.
;Figure 3b:Data show the RPLC profile of the 2X purified material along with the inhibitory activity and SDS PAGE analysis of fractions 20,21 and 22.

WO 90/09183 ! PCT/US90/00837 Flguri 4:0111:B4 LPS ( 10 ng/ml) was preincubated with varying doses of (A) purified BPI, and (B) polymyxin B, then tested for neuuophil stimulatory activity.
Results from two experiments shows inhibition of complement receptor expression on neutrophils with Standard Errors for replicate samples.
Figure 5: (a) A bar graph illustrating BPI expression on the surface of neutrophils stimulated with FMLP, TNF, and L,PS. (b) A bar graph illustrating maximal CR3 upregulation of human neutrophil cell surface expression.
Figure 6: A bar graph illustrating that BPI and polymyxin B inhibited more than 70% at time=0 of the neutrophil response to LPS.
Figure 7: A graph illustrating that BPI inhibits LPS activity on LAL assay.
Figure 8: A chromatogram showing a fractionated azurophile granule extract by cation exchange HPLC (step 1); the dotted line traces LPS inhibitory activity and the solid Iine traces protein absorbance.
Figure 9: A chromatogram showing a fractionated azurophile granule extract by cation exchange hiPLC (step); the dotted line traces LPS inhibitory activity and the solid line traces protein absorbance.
Figure 10: A chromatogram showing a fractionated azurophile granule extract by size exclusion HPLC (step 3); the dotted line traces LPS inhibitory activity and the solid line traces protein absorbance.
Figure 11: An SDS-PAGE gel of the azurophil granule extract, the precipitated extract, and fraction pools from the three chromatographic steps.

Figure 12: Analysis of purified BPI by microbore reverse phase I-iPLC
identifying a single major peak which accounts for 97% of the total protein.
Figure 13: A line graph illustrating inhibition of the neutrophil response to ng/ml LPS by BPI.
Figure 14: A line graph showing BPI directly binds to LPS.
Figure 15: A line graph showing BPI binding to immobilized LPS was inhibited bY PolYmYxin B.

Figure 16: A line graph showing that BPI binds to LPS in the presence of plasma.
Figure 17: A line graph showing BPI binds to LPS in the presence of serum.
Figure 18: A bar graph showing that BPI modulates pyrogenic response to LPS.

WO 90/09183 ' PCT/US90/00837 Detailed Descriytion of the Invention The present invention provides a method of inhibiting lipopolysaccharide (IrPS)-mediated stimulation of cells. This method comprises contacting the cells, in the presence of a cell-stimulating amount of lipopolysaccharide, with BPI
in an amount effective to inhibit cell stimulation The amount of BPI effective to inhibit cell stimulation will vary according to the conditions present. T'he amount effective to inhibit cell stimulation is preferably from about 100 ng to about 100 mg, with the most preferred amount being from about 10 ~g to about 10 mg.
Neutrophils and monocytes are the cells of greatest importance with regard to the ~ 5 application of the subject invention. However, other cells such as endothelial cells are also affected by LPS and may be used in this invention.
In the preferred embodiment purified BPI is used. BPI also comprises 20 recombinant BPI and biologically active polypeptide analogs thereof. One suitable analog 'of BPI comprises a polypeptide which has a molecular weight of about 25kD and corresponds to the h'-terminal amino acid sequence of BPI.
25 ~ used herein a biologically active polypeptide analog of Bactericidal/Permeability Increasing Protein means a polypeptide which has substantially the same amino acid sequence as, and the biological activity of, native or naturally-occurring Bactericidal/Permeability Increasing Protein.
30 ~e ~vention further provides a method of treating a gram negative bacterial infection. This method comprises contacting the bacterial infection with purified BPI nr a biologically active polypeptide analog thereof in an amount effective to inhibit LPS-mediated stimulation of cells and thereby treat the bacterial infection.
It would be clear to one skilled in the art that gram negative bacterial infection includes gram negative sepsis, the most common nosocomial infection which 5 causes death. Generally, gram negative sepsis is a severe toxic, febrile state resulting from infection with pyogenic microorganisms, with or without septicemia.
The gram negative bacterial infection may be associated with endotoxic shock or 10 an inflammatory condition. The inflammatory condition may, for example, be associated with disseminated intravascular coagulation (DIC), adult respiratory distress syndrome CARDS), or renal failure.
In the preferred embodiment the gram negative bacterial infection is present in a subject, most preferably, a human being.
Additionally, the present invention provides for a composition for treatment of a gram negative bacterial infection. This composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit LPS-mediated stimulation of cells and a suitable carrier.
In the preferred embodiment the BPI or a biologically active polypeptide analog thereof is administered in a pharmaceutically acceptable carrier.
Pharmaceutically acceptable carrier encompasses any of the standard pharmaceutical carriers such as sterile solution, tablets, coated tablets and capsules. Typically such carriers contain excipients such as starch, milk, sugar, certain types of clay, gelatin, stensic acid, talc, vegetable fats or oils, gums, glycols, or other known excipients.
Such carriers may also include flavor and color additives or other ingredients.
Compositions comprising such carriers are formulated by well known conventional WO 90/09183 , PCT/US90/00837 methods. However, the composition comprising BPI or a biologically active polypeptide analog thereof in an amount effective to suppress LPS mediated stimulation of neutrophils or monocytes is previously unknown.
In this method, the administration of the composition may be effected by any of the well known methods, including but not limited to, oral, intravenous, intramuscular, and subcutaneous administration.
In the practice of the method of this invention the amount of BPI or a biologically 0 active polypeptide analog thereof incorporated in the composition may vary widely. Methods for determining the precise amount are well known to those skilled in the art and depend inter alia upon the subject being treated, the specific pharmaceutical carrier and route of administration being employed, and the frequenry with which the composition is to be administered.
The present invention additionally provides a method of treating a subject suffering from endotoxin-related shock caused by a gram negative bacterial infection which comprises administering to the subject an amount of BPI
effective to combat the gram negative infection and treat the subject so as to alleviate the endotoxin-related shock. Endotoxins, as used herein, are substances containing Iipopolysaccharide complexes found in the cell walls of microorganisms, principally gram-negative bacteria.
Further, the invention provides a method of treating a subject suffering from disorder involving disseminated intravascular coagulation. The method comprises administering to the subject an amount of BPI effective to alleviate the symptoms of disseminated intravascular coagulation and thereby treat the subject.
As used herein, the term disseminated intravascular coagulation is a complex disorder of the clotting mechanisms, in which coagulation factors are consumed at an accelerated rate, with generalized fibrin deposition and thrombosis, hemorrhages, and further depletion of the coagulation factors. Moreover, disseminated intravascular coagulation may be acute or chronic.
Further, the present invention provides a method of treating a subject suffering from endotoxemia caused by a gram negative infection which comprises administering to the subject an amount of BPI effective to combat the gram negative bacterial infection and neat the subject suffering from endotoxemia.
0 As used herein endotoxemia means a condition in which the blood contains poisonous products, either those produced by the body cells or those resulting from microorganisms, i.e. gram negative bacteria.
The invention also provides a method of treating a subject suffering from endotoxin-related anemia caused by a gram negative bacterial infection which comprises administering to the subject an amount of BPI effective to combat the gram negative bacterial infection and treat the subject so as to alleviate endotoxin-related anemia.
The present invention further provides a method of treating a subject suffering from endotoxin-related leukopenia caused by a gram negative bacterial infection.
The method comprises administering to the subject an amount of BPI effective to 26 combat the gram negative bacterial infection and treat the subject so as to alleviate endotoxin-related leukopenia. Further, the invention includes a method of treating a subject suffering from endotoxin-related thrombocytopenia caused by a gram negative bacterial infection which comprises administering to the subject an amount of BPI effective to combat the gram negative bacterial infection and xreat the subject so as to alleviate endotoxin-related thrombocytopenia.

, WO 90/09183 ' PCT/US90/00837 The invention also provides a method of inhibiting a pyrogen which comprises contacting the pyrogen with an amount of BPI so as to inhibit the pyrogen.
As used herein, a pyrogen is any fever-producing substance; exogenous pyrogens include bacterial endotoxins, especially of gram-negative bacteria; endogenous pyrogen is a thermolabile protein derived from such cells as mononuclear leukocytes which acts on the brain centers to produce fever.
Furthermore, the invention also includes a method of inhibiting lipopolysaccharide-mediated tumor necrosis factor production by cells which p comprises contacting the cells in the presence of a cell-stimulating amount of lipopolysaccharide, with BPI in an amount effective to inhibit lipopolysaccharide-mediated tumor necrosis factor production by cells.
~ 5 The invention also includes a method of inhibiting gram negative bacteria-mediated tumor necrosis factor production by cells which comprises contacting the gram negative bacteria with BPI in an amount effective to inhibit gram negative-mediated tumor necrosis factor production by cells.
20 The amount of BPI effective to inhibit cell stimulation will vary according to the conditions present. The amount effective to inhibit cell stimulation is preferably from about 100 ng to about 100 mg, with the most preferred amount being from about l0ug to about 10 mg.
In the above-described methods, the BPI comprises recombinant BPI or a biologically active polypeptide analog thereof. Moreover, the biologically active polypeptide analog of BPI comprises a polypeptide which has a molecular weight ,of about 25 kD and corresponds to the N-terminal amino acid sequence of BPI.
Moreover, the invention includes a composition for the treatment of a subject suffe:-ing from endotoxin-related shock. The composition comprises a purified BPI
or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related shock and a suitable carrier.
Further, the invention includes a composition for the treatment of a subject suffering from disseminated intravascular coagulation. The composition comprises a purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from disseminated intravascular coagulation and a suitable carrier.

The invention also includes a composition for the treatment of a subject suffering from endotoxemia comprising a purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxemia and a suitable carrier.
The invention additionally provides a composition for the t~eaunent of a subject suffering from endotoxin-related anemia comprising purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related anemia and a suitable carrier.
Additionally, the invention provides a composition for the treatment of a subject suffering from endotoxin-related leukopenia. The composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related leukopenia and a suitable carrier.
Further provided is a composition for the treatment of a subject suffering from endotoxin-related thrombocytopenia. The composition comprises purified BPI or a biologically acti~re polypeptide analog thereof in an amount effective to treat a subject suffering from endotoxin-related thrombocytopenia and a suitable carrier.

WO 90/09183 t PCT/US90/00837 Also, the invention provides a composition for inhibiting lipopolysaccharide-mediated tumor necrosis factor production by cells comprising purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit lipopolysaccharide-mediated tumor necrosis factor production by cells and a suitable carrier. Moreover, the invention provides a composition for inhibiting 5 gram negative bacteria-mediated tumor necrosis factor production by cells comprising purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit gram negative bacteria-mediated tumor necrosis factor production by cells and a suitable carrier.

The invention provides a composition for inhibiting a pyrogen. The composition comprises purified BPI or a biologically active polypeptide analog thereof in an amount effective to inhibit a pyrogen.
Additionally, the invention provides a method of preventing a condition associated with a gram negative bacterial infection in a subject which comprises administering to the subject an amount of BactericidaZ/Permeability Increasing Protein effective to prevent the gram negative bacterial infection and thereby prevent the condition.
In the previously described method the condition is any of the conditions selected from the group consisting of endotoxin-related shock, endotoxemia, endotoxin-related anemia, endotoxin-related leukopenia, or endotoxin-related thrombocytopenia Moreover, also provided is a method of preventing a disorder involving disseminated intravascular coagulation in a subject which comprises administering to the subject an amount of Bactericidal/Permeability Increasing Protein effective to prevent the symptoms of disseminated intravascular coagulation and thereby preventing the disorder.

Finally, the invention provides a method of isolating and recovering purified Bactericidal/Permeability Increasing Protein which comprises: (a) obtaining a crude sample of Bactericidal/Permeability Increasing Protein; and (b) separating the crude sample by column chromatography using de-pyrogenated solutions thereby isolating and recovering purified Bactericidal/Permeability Increasing Protein. Moreover, in this method the Bactericidal/Permeability Increasing Protein comprises native Bactericidal/Permeability Increasing Protein or a biologically active polypeptide analog thereof. Further, the biologically active polypeptide analog of Bactericidal/Permeability Increasing Protein comprises a polypeptide which has a molecular weight of about 25 kD and corresponds to the N-terminal amino acid sequence of Bartericidal/Permeability Increasing Protein.
We have found that biological activity level of BPI varies depending on the method used for obtaining BPI. It appears that depyrogenated BPI, i.e. BPI
isolated and recovered by the above-described method using de-pyrogenated ~ 5 solutions, shows a much higher level of biological activity than pyrogen-containing BPI (Table 5 and Figure 4A).
This invention is illustrated in the Experimental Details and Results sections 2o which follow. These sections are set forth to aid in an understanding of the invention but are not intended to, and should not be construed to, limit in any way the invention as set forth in the claims which follow.

~ WO 90/09183 ' PCT/US90/00837 Examyie 1 Materials and Methods:
t 0 Lipopolysaccharide from E.E. Coli 0111:B4, S. Syphimurium wild type, glycolipid from S~tvnhimurium RE mutant, and Lipid A from ~y~7himurium RE mutant, and LPS from P. aeru '~tnosa were purchased from RIBI Immmunochem Research, Inc., Hamilton, MT; Fmet-Leu-Phe (FMLP) and polymyxin B Sulfate from Sigma Chemical Co., St. Louis, MO; Hank's Balanced Salt Solution without calcium, magnesium and phenol red (HBSS) from Hazelton Research Products, Denver, PA; Ficoll-Paque* Percoll~and MacrodeX from Pharmacia Inc., Piscataway, NJ;
TNF and anti-TNF from Endogen, Boston MA; Fluorescein conjugated goat-anti-mouse IgG from TAGO Inc., Burlingame, CA: IgGl control antibody from Coulter Immunology, Hialeah, FL; Phycoerythrin (PE) conjugated anti CR3 (Leu-15) and IgG2a control from Becton Dickinson, Mountain View, CA, Anti CR1 monoclonal antibody, Yz-1, was a kind gift from Dr. Rick Jack at Harvard University Granulocytes were isolated from buffy coats obtained from local blood banks.
Buffy coats were. diluted 3-4X in HBSS and granulocytes were separated from 'mononuclear cells by centrifugation through 64% Percoll. The pellet was subjected to diisopropylfluorophosphate (DFP), washed, and resuspended in ice *Trademark cold lysis buffer ( 10 mM PIPES, pH 6.8, 100 mM KCL, 3mM NaCI, 3.5 mM
MgCl2) aad disrupted by nitrogen cavitation (Purr Instrument Co., Moline, IL).
Azurophil granules were isolated on discontinuous Percoll gradients as described by Borregaard. [Borregaard, N., J.M. Heiple, E.R. Simons, and R.A. Clark, J.
Cell.
Biol., 97: 52-61 (1983)] The azurophil granules were collected and Percoll was removed by centrifugation at 180,000 X G for 2 hours. The granules were lysed by 4 cycles of freeze-thaw followed by 1 minute of sonication. The lysed granules were extracted in an equal volume of 100 mM glyrine, pH 2 by vortexing intermittently for 1 hour at room temperature. The acid extract was clarified by centrifugation at 30,000 X G for 20 minutes and at 200,000 X G for 30 minutes.
Venous blood was drawn from healthy volunteer donors into acid citrate dextrose ~ 5 anticoagulant and immediately placed on ice. Five parts of blood were mixed with 1 part of cold Macrodex;' and allowed to settle for 1.5 - 2 hours at 4 ~ C.
Leukocyte-rich plasma was washed 1X in cold HBSS, then resuspended in HBSS
and layered over Ficoll-Paque. ~ If significant erythrocyte contamination was 2o present, the granulocyte pellet was subjected to hypotonic lysis. The cells were washed 2X in HBSS and resuspended in HBSS + 2% autologous plasma to give a final granulocyte concentration of 1 X 1f6/ml in the incubation mixture.
25 ~ BPI Punfication Approximately 2 mg of crude azurophil granule extract was separated by size on a Biosil ('1"SK-250) (7.8 mm x 600 mm) high performance size exclusion column using 50 mM glycine and 100 mM NaCl buffer, pH 2.0, under isocratic conditions of a flow rate of 1 ml/min. Column fractions with the greatest LPS inhibitory activity contained a large proportion of the 54 KD species as shown by SDS
*Trademark . WO 90/09183 ~ PCT/US90/00837 PAGE. These TSK fractions were pooled and run over an Aquapore weak cation exchange (WCX) column (2.1 mm X 30 mm) using 50 mM citrate, pH 5.5, and eluted in a gradient of 0-75%, of 50 mM citrate and 1 M NaCI (Buffer B) in 25 min, then 75-100% Buffer B in 5 min with a flow rate of 200 ml/min. Material of 57 KD was recovered from canon exchange and appeared as a single band on SDS page. In some experiments BPI was further purified by reverse phase HPLC
on a Vydac C4 column loaded for 12 min in 0.1% CH3CN plus 0.1% TFA, in 30 min with a flow rate of 200 ml/min (Raisin Instruments, Emeryville, CA).
0 Neutrophil stimulation Isolated neutrophils were kept on ice until incubated with and without stimuli at 37' C for 30 minutes. Following the incubation, cells were washed in a large volume of cold PBS + 0.05% Na Azide + 2% autologous plasma. Pellets were t5 divided in two, one stained with SO~uI control IgGl antibody (20~ug/1x106 cells), the other with 50 ul of 20 pug/ 1x106 cells anti-CR 1 for 30 minutes at 0' C.
Following this incubation the cells were washed 2X with PBS + autologous plasma, then stained with goat-anti-mouse IgG-FTTC, and in some experiments, 20 20u1 of IgGZa-phycoerythrin (PE) in control wells, and 20A1 Leu-15 PE in test wells. Following a 30 minute incubation at 0' C and 2 more washes, the cells were analyzed by flow cytometry on a Becton Dickinson FACStar flow cytometer (Becton Dickinson, Mountain View, CA). Neutrophil stimulation was measured 25 by comparing mean fluorescence intensity of samples which had been incubated in HBSS + 2% autologous plasma alone (control) to those incubated with LPS
or LPS which had been preincubated for 30 minutes at 37' C with BPI or polymyxin B. Data are expressed as % stimulation or % inhibition and were calculated using the mean fluorescence intensity (Fl), on a log scale, according to:
*Trademark % Stimulation = [(Experimental - Control)/ (Maximum - control)] X

and % Inhibition= 1-[(+Inhibitor)-(Control)]/
[(-Inhibitor)-(Control)] X 100.

Vapor phase hydrolysis of BPI and amino acid derivitization was performed using a Pico-tag Workstation (Waters, Milford MA) and chromatographic analysis of the phenylthiocarbamyl amino acids was performed on an applied Biosystems 130 A
MPL,C using Protocols provided by the manufacturer.
BPI N-terminal sequence was analyzed by automated Edman degradation using an applied Biosystems 477A pulse liquid phase sequenator (Applied Biosystems, Foster city, CA). Phenyltheohydantion amino acid analysis was performed on line using an applied biosystems Model 120A liquid chromatograph.
Human neutrophils may be stimulated both in vivo and in vitro by lipopolysaccha,ride. Upon activation, surface expression of receptors for C3b and C3bi (CR1 and CR3 respectively), increases. Using the Fluorescence Activated Cell Sorter (FACS), fluorescence intensity of freshly isolated human neutrophils was measured following stimulation with increasing doses of 0111:84 LPS
(Figure la). Because commonly observed maximum stimulation was at or above 10 ng/ml, experiments testing for inhibition of 0111:84 LPS used 10 ng/ml as the *Trademark WO 90/09183 ~ PCT/US90/00837 stimulatory dose. All experiments were performed in duplicate. In most experiments, data is shown only for CR1 since we did not observe any condition where neutrophil stimulation caused upregulation of CR1 or CR3 alone (M.
Marra et al. ( 1990) J. Immunol. L(2):662-666).
To determine whether proteins found in neuuophil azurophil granules could interfere with the neutrophil response to LPS, crude acid extracts of azurophil granules were pre-incubated with LPS for 30 minutes at 37' C. The mixture was then tested for its ability to stimulate neuuophils. Azurophil protein (l,ug/ml) could effectively block stimulation of 1 X 106 polymorphonuclear leukocytes (PMN)/ml by 10 ng/ml of LPS (Figure lb). This effect was not observed using glycine extraction buffer preincubated with LPS, nor was there any stimulation of neutrophils using crude extract or glycine buffer control (data not shown).
To further investigate which of the proteins in the extract was/were responsible for inhibitory effect, crude acid extracts were separated by reverse phase HPLC;
each peak was assayed separately for LPS inhibitory activity. The identity of each of the peaks was previously determined using a two-dimensional purification 2~ approach involving microbore reverse phase HPLC in first dimension followed by SDS PAGE, electroblotting and microsequencing. The azurophil proteins can be ZUROPHIL GRA NULE
DERIVED

A -Peak Identity PROTEINS

1 Defensins CYCRIPACIAGERRY

(HNP-2) 2 Granulocidin VCSCRLVF~CRRTGLR

(HNP-4) 3 Eosinophil Cationic Protein (ECP) XPPQFTRAQWFAIQH

4a Eosinophil-Derived Neurotoxin ~5 (EDN) KPPQFTXAQXFETQX

4b CathepsinG IIGGRESRPHSRPYM

Sa Lysozyme KVFERXELARTLKRL

2~ Sb Eosinophil Major Protein (MBP) TCRYLLVRSLQTFSQ

6 Unlmown IVGGRKARPXQFPFL

7 Unknown IVGGHEAQPHSRPYM

8a MyeloperoxidaseV N C E T S C V Q Q P P
C F P

8b Elastase IVGGRRARPHAXPFM

9 Bactericidal/Permeability Increasing Protein (BPI) VNPGVVVRISQKGLD

~ WO 90/09183 PCT/US90/00837 resolved into 10 discrete peaks whose identities are shown in Table 1. The amino acid sequences shown are for the first 15 amino acids of the N-terminal.
LPS inhibitory activity of l,ug of each peak is shown in Figure 2. As shown, peak 9 had the highest LPS neutralizing activity. The major protein species in this peak has N-terminal identity with Bactericidal/Permeability Increasing Protein (BPI) described previously (Weiss, J., P. Elsbach, I. Olsson and H. Odeberg, J.
Biol.
Chem, 253 (8): 2664-2672 (1978)). BPI has been shown to contain the majority of the gram negative bactericidal activity in azurophil granule protein extracts.
0 Cathepsin G showed some inhibition of LPS, but the data between experiments were not as reproducible as for peak 9. Cathepsin G has been shown to bind to LPS in vitro and to kill gram negative organisms, although to a lesser extent than BPI. Other proteins which have demonstrated microbicidal activity against gram negative organisms are elastase and the defensins. However, these proteins ( 1 iug/ml) did not inhibit the stimulatory activity of LPS on neutrophils.
LPS inhibitory activity of crude azurophil extracts was further characterized and purified using size exclusion and ion exchange followed by reverse phase chromatography. LPS inhibitory activity comigrates with a pure 57 ICD band seen on SDS PAGE (Figure 3b).
Because the buffer used in the RPLC separations [CH3CN and 0.1 plc trifluoroacetic acid ('IFA)] significantly diminishes the LPS inhibitory activity of BPI (data not shown), and since the material purified from ion exchange chromatography was of high purity as judged by SDS PAGE, size exclusion/ion exchange material was used to generate a dose response curve (Figure 4a). Data is shown from two experiments, each performed in duplicate. This size exclusion/ion exchange purified material was confirmed to be BPI by N-terminal sequence analysis. Protein concentration was determined by amino acid analysis.

As s< en in Ftgnre 4a, about 90 ng/ml of BPI is required for maximal inhibition of tse neutrophil response to 10 ng/ml 0111/84 LPS. The neutrophil response to formulated peptide ( 10''M FMLP) was not inhibited by BPI (data not shown).
Figure 4b shows a similar does response curve for the polypeptide antibiotic Polymyxin B (PMB). Polymyxin B binds to the Lipid A moiety of LPS and neutralizes some of its toxic effects both inin vivo and in vitro. polymyxin B
has been demonsuated to bind to LPS stoichiometrically (Morrison, D.C. and D.M.
Jacobs, Immunochem, 13: 813-818 (1976)). The calculated amount of PMB
required to inhibit 10 ng/ml of smooth LPS is approximately 0.67 nM. In the subject experiments 0.4 ng/ml, or 0.36 nM of polymyxin B was required to completely inhibit neutrophil stimulation using 10 ng/ml of LPS. 90 ng/ml, or 1.58 nM BPI was required for 100% inhibition of 10 ng/ml LPS.
Therefore, on a molar basis the amount of BPI required to inhibit LPS
stimulation of neutrophils in vitro was approximately 4X the amount required for polymyxin B.
To test whether BPI can inhibit LPS from other gram negative organisms, LPS
2p molecules with varying polysaccharide chain lengths and Lipid A were tested in the subject system against 90 ng/ml of 2X purified BPI. Data shown in Table 2 demonstrates that although the stimulatory dose may vary between these molecules, LPS from both smooth and rough chemotypes as well as Lipid A are all inhibited by BPI.

WO 90/09183 ' PCT/US90/00837 10 NG f ML 1 NG~ML
5 E.COLIO111:B4 97 ' S. TYPHIMURIUM

S. TYPHIMURIUM

LIPID A
~ 5 P. AERUGINOSA 112 ' ' Low to no stimulation tration at this endotoxin concen As previously discussed, Bacterial/Permeability-Increasing protein (BPI) is a cationic SO-60,000 m.w. protein first purified from human neutrophil granules by Weiss et al. (Weirs, J., P. Elsbach, I. Olsson and H. Odegerg. 1978.1. Biol.
Chem.
253:2664.). BPI alters bacterial cell membrane permeability and has bactericidal activity specifically against gram negative organisms. To date, the literature on BPI has focused exclusively on its bactericidal activity.
We report that BPI binds to LPS and inhibits both neutrophil and monocyte responses to soluble LPS in vitro. BPI also inhibits LPS activity in the Limulus ~ebocyte Lysate assay. Our research has identified BPI as a lead molecule for the development of novel therapies against endotoxic shock.
In response to LPS, human neutrophils upregulate cell surface expression of 2~ complement receptors CR1 and CR3 (Figures la and Sb). To measure this neutrophil response to LPS, we incubated freshly isolated human neutrophils with E. Coli 0111:B4 LPS (Figure 4a), and showed that maximal CR1 upregulation is observed using 10 ng/ml LPS (Figure 4). Neutrophil stimulation with LPS was not .inhibited by exogenous anti-TNF antibodies, suggesting that LPS acted directly on neutrophils in this system.
BPI inhibits the neutrophil response to LPS (Figure 4a). Inhibition of CR
upregulation was complete at a dose of approximately 1.8-3.6 nM (100-200 ng/ml) ; BpI compared to 0.4 nM polymyxin B required to inhibit 10 ng/ml smooth LPS
(approximate m.w. 15,000) is about 0.7 nM, matching closely with the observed ~ WO 90/09183 ~ PCT/US90/00837 value of 0.4 aM. On a molar basis, the amount of BPI requires ibit LPS
was approximately 5-fold greater than the amount required for poly ..:pcin B.
BPI inhibits LPS-mediated neutrophil stimulation but not stimulation by either FMLP or TNF (Table 3). These data demonstrate that BPI inhibits LPS directly and does not disrupt neutrophil mechanisms involved in CR upregulation.
Neutralization of LPS by BPI occurred rapidly. Even without preincubation, both BPI (and polymyxin B) inhibited more than 70% of the neutrophil response to 0 LPS (Figure 6). Maximal inhibition was seen following only 5 minutes of preincubation.
BPI inhibits CR upregulation stimulated by LPS from smooth and rough bacterial strains, as well as lipid A (Table 4) Because of the broad range of BPI
activity against these different forms of LPS, among which only lipid A and 2-keto-3-deoxy-octonate are shared determinants, it is likely that LPS inhibition by BPI is affected through lipid A.
gpI inhibits other LPS-mediated activities. At a concentration of approximately 9 nM, or 500 ng/ml, BPI significantly inhibited LPS activity in the LAL assay (Figure 7). When LPS and BPI were added together without preincubation no inhibition was observed (data not shown), indicating that BPI acted on LPS, and had no effect on the LAL assay system. BPI also inhibits LPS-mediated TNF
production by human adherent mononuclear cells (Table 5).
.

Effect of BPI on Neutrophil Stimulation by Various Agents Inhibition of CR Upregulation on Neutrophils Stimulus Dose % Inhibition % Inhibition CRl CR3 LPS 10 ng/ml ~ 109 102 FMLP 10'' M 9 11 rTNF 50 U/ml 0 0 Neutrophils were incubated with E.E. Coli O111:B4 LPS, FMLP or TNF
preincubated in the presence or absence of 2.7 nM BPI. Data is reported as percent inhibition of CR expression in response to each stimulus preincubated with buffer alone.

WO 90/09183 ' PCT/US90/00837 Inhibition of LPS and Lipid A induced Neutrophil Stimulation by BPI

Inhibition of CR Upregulation on Neutrophils Stimulus Dose CRl CR3 (ng/ml) 96 Inhibition % Inhibition None - 0 0 E.E. Coli 011:84 LPS 10 100 99 S~yphimurium Wild Type LPS 10 104 100 S. xryhi RE Mutant LPS 1 97 95 ~ 5 S. yhimurium RE Mutant Lipid A 1 111 104 Neutrophils were stimulated with LPS and lipid A preincubated with and without 2.7 nM purified BPI. Results are expressed as percent inhibition of fluorescence intensity observed with each type of LPS alone.

BPI Inhibits LPS-Induced TNF Production by Human Monocytes TNF (nQ/mll Produced in Resuonse to LPS Preincubated With':
LPS Medium 100 ng/ml 500 ng/ml 250 ng/ml Buffer (ng,~ml) ~ one Polymy~dn B BPI $EI n r 0.1 61 0 0 0 81 s~. Coli 0111:B4 LPS, was preincubated with BPI or polymyxin B (Pr~~l, than added to adherent peripheral blood mononuclear cells. TNF produc::un was assayed by ELISA.

WO 90/09183 ' PCT/US90/00837 BPI was first purified by Elsbach and Weiss in 1978. In our initial studies we isolated BPI from azurophil granule extracts in a single step by reverse phase HPLC. Recovery of BPI activity from reverse phase was poor, probably due to the denaturing conditions. Here we show the purification of LPS inhibitory activity using only non-denaturing steps and demonstrate that most of the activity from neutrophils comigrates with BPI. Improvements in the purification have also led to very high specific activity material as will be shown in the following section.
Figures 8-10 show the three chromatographic steps currently employed in our lab.
The absorbance is traced by the solid line and LPS inhibitory activity on the dotted lines. Table 6 shows the recovery of activity and protein and the specific activity, as measured in~arbitrary LPS neutralizing units (NU). One neutralizing unit is that amount of BPI that inhibits OS E.U. LPS by SO% in the LAL test. A
Commassie stained SDS-PAGE gel, of these pools is shown in Figure 11.
Analysis of the purified BPI by microbore reverse phase HPLC (Figure 12) identified a single major peak which accounted for 97% of the total protein by integration. Tryptic mapping of . BPI allowed us to sequence several major fragments which further confirm the identity of the protein. The full length published sequence for BPI is known (P.W. Gray et al. (1989) J. Biol. Chem.
~4( 16):9505).

II. PURIFICATION OF BPI UNDER RIGOROUSLY PYROGEN-FREE
CONDITIONS
Materials and Methods eaeents USP grade sterile irrigation water was obtained from Travenol Laboratories Inc., Dee~eld IL; Pyrosart filters from Sartorius GmbH, W.
Germany; CM Sepharose FF from Phasmacia, Upsaza, Sweden; Polyaspartamide weak ration exchange HPLC column (100 X 4.6mm) from the Nest Group, Southborough MA; Glycine and Bio-Sil 6250 size exclusion HPLC column (600 X 7.5mm) from Bio-Rad Laboratories, Richmond CA; Polyacrylamide electrophoresis gels from Novex, Encitas CA; Sequencing and amino acid analysis reagents and buffers from Applied Biosystems Inc., Foster City, CA;
Trifluoroacetic acid, constant boiling HCL, hydrolyzate amino acid standard, and ~ 5 BCA protein assay reagents from Peirce Chemical Co., Rockford, IL; Limulus Amebocyte Lysate assay from Whittaker Bioprodurts, Inc., Walkersville, MD;
Lipopolysaccharide from RIBI Immunochem Research, Inc., Hamilton, MT;
HPLC grade Acetonitrile from J.T. Baker, Phillipsburg, NJ; all other buffers and salts used were reagent grade. 18 megohm purity water was prepared by Lab Five ultrapure water system from Technic, Seattle, WA. O.SM NaOH for sanitization was prepared from reagent grade NaOH pellets and USP water.
(~ranuls extracts from Neutroo~ were prepared as described (U.S. Serial No.
199,206, filed May 26, 1988) except that the percoll separation of azurophil granules was omitted. Instead, whole granule fractions were obtained by centrifuging the post nuclear supernatant at 17,000 g for 20 minutes. The granule pellet was then suspended in a volume of 1 ml of 50 mM glycine pH 2 for every ;4X10E8 cells lysed. Resuspended granules were lysed by five freeze/thaw cycles on dry ice ethanol followed by vigorous agitation for one hour at 4 degrees C.
*Trademark . . WO 90/09183 ' PCT/US90/00837 The soluble extract was obtained by centrifugation at 30,OOOg for 30 minutes.
Limulus Ameboc~r~ysat~ was performed as directed by the manufacturer.
Where necessary the pH of samples was adjusted to neutrality by the addition of pyrogen free O.SM phosphate buffer pH 7.4. and salinity was decreased to < 150 mM by dilution with USP water.
LPS neutralization assay: was performed as previously described (M. Marra et al.
( 1990) J. Immunol. x(2):662-666).
High salt fractionation of granule extracts: 200 mgs of extracted protein were pooled from various preparations and kept on ice. 1 volume of sterile SM NaCI
was added for every 4 volumes of extract. The resulting precipitate was pelleted by centrifugation at 20,000 g for 20 minutes at 4 ~ C. This supernatant was prepared for CM sepharose chromatography by diluting with 4 volumes of USP
irrigation water and adjusting the pH with enough 1M Tris pH 7.4 to give a final concentration of 50 mM. Only fresh, sterile, pyrogen free stock salts and buffers were used.
CM Se~harose chromatoQr_aRhv: An XK-16 column (Pharmacia) was packed with sufficient resin to give a bed volume of Smls. The column was installed on a gradient FPLC equipped with a Pl pump for sample loading. Prior to use, all s~~s in contact with the mobile phase were extensively washed with O.SM
NaOH. The column was sanitized by washing at 0.2 mls/min. with O.SM NaOH
for 4 hrs. The column was then re-equilibrated and a blank run was performed.
Fractions from the blank run and eluenu were tested by LAL assay for pyrogenicity. Prepared exuact was loaded at a flow rate of 400 mls/hr. Once ~ loaded the column was washed with 2 to 3 column volumes of starting buffer.
The granule extract was kept on ice during loading. The column was run at room tcm~~erature.
Weak canon exchange HPLC: was performed using an Eldex ternary gradient pump equipped with a Rheodyne injector and a Gilson model 1118 U.V. detector.
Wettable surfaces were washed with O.SM NaOH followed by extensive rinsing with USP water to remove all traces of base prior to installing the column.
Blank fractions and eluents were tested for pyrogenicity as above.
C'~el ~rmeation HPLC: was performed with the same precautions and equipment 0 outlined for weak cation exchange HPLC.
Poly~lamide gel electrophoresis: 8 to 16% acrylamide gradient gels were purchased from Novex and run according to the manufacturers specifications.
Protein seguence determination: An Applied Biosystems 477A pulsed liquid phase sequenator equipped with a 120A PTH amino acid analyzer was used for automated edmund degradation.
Microbore reverse phase HPLC: Material for protein sequencing was prepared by desalting on a 30 X 2.1 mm Aquapore butyl column. The gradient used was to 100% B in 30 minutes at a flow rate of 200 ml/minute. Detector settings were 214 nm wavelength at 2.0 absorbance units full scale (see insert figure X).
25 pn HP 3396A was used to integrate and plot data.
,; was performed on the system described above using the PTC column, buffers and separation conditions provided by ABI. Sample 30 hydrolysis and PTC derivatives were prepared using a Pico-Tag workstation from the Waters chromatography division of Millipore using the manufacturer's protocols.
*Trademark WO 90/09183 , PCT/US90/00837 Protein concentrations were determined using .A method instructions 23230, 23225 from Peirce Chemical Co. In order to minimize buffer interference, samples were diluted 1(I fold and the micro reagent protocol was used.
RESULTS
BPI purified from azurophil granules was previously shown to inhibit neutrophil activation by LPS and to inhibit LP;S directly in the LAL assay. In order to 10 ~rther define the role of BPI and investigate the presence of other similar molecules in both azurophil and specific granules, we undertook the purification of LPS inhibitory activity from whole ;granules extracted at acid pH.
Preliminary studies verified the presence of LPS inhibitory activity in the crude extract.
To identify the endotoxin neutralizing activity we attempted its purification from whole granule extracts. Purification. of LPS neutralizing activity was greatly enhanced by the observation that high concentrations of NaCI ( 1M) caused the reversible precipitation of about ninety percent of the protein present in the granule extract. Essentially all of the LPS inhibitory activity remained in the soluble supernatant. The soluble fra~aion was then diluted, to reduce the ionic strength, and further purified and con<:entrated by CM sepharose cation exchange chromatography. A broad peak of activity eluted which was subsequently further ~ p~~d ~~g a polyaspartamide high performance cation exchange column. A
somewhat sharper peak of activity was recovered which comigrated with a major protein of about 55,000 molecular weight by SDS-PAGE along with several lower molecular weight proteins. Gel permeation HPLC was used as the final purification step and identified a peak of activity which eluted with a single sharp protein peak. The purified protein migrated as two closely spaced bands on SDS-PAGE at 55,000 molecular weight. 25% of the total endotoxin neutralizing activity was recovered with a 250 fold purification.
The purified endotoxin neutralizing protein was subjected to reverse phase HPL,C
followed by N-terminal sequence analysis by automated Edman degradation. The sequence, shown in figure 6 was identified as bacterial permeability increasing protein by virtue of complete homology through 39 residues. In addition the amino acid composition of the purified molecule was virtually identical to that of BPI (data not shown).
To investigate whether both closely spaced bands were BPI we subjected the purified proteins to western blotting analysis using BPI-specific rabbit polyclonal antisera raised against a synthetic peptide comprising amino acids 1-20 of BPI.
Both bands were immunoreactive. The differences may arise from glycoslyation.

. , WO 90/09183 ' PC1"/US90/00837 Purification of BPI under rigorously pyrogen-free conditions, as described in section II resulted in a more potent BPI preparation as shown by the dose Recovery Specific Acts Acts Extract 100% 100% 0.11 NU/ag Precipitated 149% 17.3% 0.93 NU/ug Step 1 35% 1.50 2.51 NU/,ug Step 2 14% 0.75 1.97 NU,Gtlg Step 3 18% 0.10% 18.9 NU/,ag response curve in Figure 13. Inhibition of LPS-mediated CR upregulation was complete at 25 ag/mI BPI, representing a 4-fold increase in activity compared to the material used in section I. On a molar basis this BPI preparation inhibited LPS at approximately stoichiometric proportions, equivalent to molar inhibitory concentrations of polymyxin B. BPI also inhibited LPS-mediated TNF production by human adherent mononuclear cells at a lower concentration following purification under pyrogen-free conditions (Tables 7 and 8).
BPI binds to LPS (Figure 14). In these experiments, 4 ug of LPS/well was immobilized on 96 well plastic plates, then incubated with varying concentrations of BPI, and developed with anti-BPI polyclonal antisera. BPI binding to LPS
was inhibited by polymyxin B (Figure 15), demonstrating specificity of BPI
binding.
BPI binds to LPS in the presence of both plasma (Figure 16) and serum (Figure 17), demonstrating potential inin vivo efficary of BPI.

W090~09183 CA 02323630 2000-10-26 P~1'~LrS94~0083i BPI Inhibits LPS-Induced TNF
Ptoduaion by Human Monocyces T'.v'F lctJml l Ptoduccduottae to LPS Pteincubated in Res with':

LPS Medium 100 ns/ml 400 ng/mlLSO ng/ml 25 ng/ml Buffer ng/ml alone PMB BPI BPI BPI Control 0.1 48 79 0 0 0 269 1 1150 120'1 0 0 0 1292 li 0111:84 LPS, was preincubated with BPI or poiymyun B (PMH), then added to adherent petiphenl blood mononuclear cells.

T1YF produa~on was assayed by ELISA.

WO 90/09183 ( PCT/US90/00837 INHIBCIZON OP LPS-INDUCED

TNF PRODUCIION BY HUMAN MONO(.YTES

'I~f' /ml) ProduccdResponse to LPS Preincubaced (nf; in with':

1000 ng/ml100 a8/ml 250 ngJml 50 n8/ml lOng/mlBuffer LPS PolymyxinPolymycn H BPl BPI HP( Control B

10 333 18 601 257 270 f 23 270 67 436 6g7 3g 37 100 769101 114073 8343p 6g6+g4 1003+50892+47 1000 844 144 1016 20 1130=10 T78 189 1025=71773 +

.S aureus 1685;121 1541 ~ 397 1538~ 139 1268 ~ 374 1554 ~ 324 1423~447 'BPI or polymyxin B sulfate were preincubated with 0-10 ng/ml E. Coli 0111:B4 LPS or 0.1%w/v killed S.S. aureus then added adherent peripheral blood mononuclear cells. TNF production was assayed by ELISA.

5 Stage IA- ogenici~y of Gl3rcine Buffer:
305u1 of Glycine Buffer control (Supplied by Redwood City) was diluted to 7 ml in PBS (Redwood City) and mixed in polypropylene tubes (pyrogen-free). The 10 rube was labeled with notebook # 1990 and tested in a three rabbit USP
Rabbit Pyrogen assay at a dose of 2 ml/rabbit (actual injection dose was 2.1 ml/rabbit).
The product was non-pyrogenic; it produced a total temperature rise for all three rabbit of 0.4 C.
304 ul of BPI (Lot 78038, dated 8/19/89) was diluted to 7 ml using PBS
(Redwood City) and mixed in polypropylene tubes (pyrogen-free). The Tube was labeled with notebook #20170 and tested in a three rabbit USP Pyrogen assay at a dose of 2.0 ml/rabbit.
Th° Product was non-pyrogenic as demonstrated by a total temperature rise of 02' C.
~Ee II- ,~oEenici of BPI ire-incubated with endotoxin:
; Endotoxin from E. Coli OSS.BS (Sigma Chemicals) was diluted in PBS (Redwood City) to 4096 EU/ml. This concentration was confirmed by the T .AI - Assay.

9~V0 90/09183 PCT/US90/00837 304 ul of BPI (Lot 78038, dated 8/19/89) was diluted to 7 ml with the PBS
diluted endotoxin (4096 EU/ml) hereinabove using polypropylene tubes. The tube was mixed by vortexing to effect mixing. The BPI+Endotoxin and Endotoxin in PBS
were incubated at 37' C in a water bath for 30 minutes. Following incubation at 37' C the BPI + Endotoxin showed an endotoxin concentration of 122 EU/ml. The endotoxin diluted in PBS did not show a change in the end point of 4096 EU/ml.
The BPI+Endotoxin and Endotoxin in PBS in PBS were tested in the three rabbit USP pyrogen assay and were found pyrogenic with total temperature rises of 4.6' C and 7.5' C, respectively.
To achieve improved results with the manipulations of the endotoxin preparation we switched from the E.E. coli OSS:BS from Sigma to the Official FDA
References.
A vial of EC-5 was rehydrated with PBS (Redwood City) to 2 ml to give a concentration of 5000 EU/ml. We verified by the label claim of 10,000 EU/ml by LAL assay.
The BPI+Endotoxin sample was prepared by adding 38,u1 of PBI (Lot 78038) to 73 ml of PBS plus 320u1 of the 5000 EU/ml of EC-5 endotoxin. The preparation v~ wed in a polypropylene tube(pyrogen-freed) and mixed well. An 8.0 ml sample of EC-5 endotoxin was prepared in PBS(Redwood City) to the same concentration without the addition of BPI. Both samples was incubated at 37' C
for 30 minutes in a water bath.
The two samples were tested for endotoxin activity using the LAL assay. The BPI + Endotoxin was negative. The endotoxin sample was positive at the target of 2~'J EU/ml (Figure 18).
Both samples were tested in the three rabbit USP Pyrogen Assay at a dose of 2.0 ml/rabbit.
The BPI+Endotoxin was non-pyrogenic and caused a total temperature rise of 1.1' C. The EC-5 endotoxin in PBS was pyrogenic and caused a total temperature rise of 3.9 ' C.

Claims (7)

1. A method of inhibiting lipopolysaccharide-mediated stimulation of cells which comprises contacting the cells in the presence of a cell-stimulating amount of lipopolysaccharide, with Bactericidal/Permeability Increasing Protein in an amount effective to inhibit cell stimulation.

2. A method of claim 1, wherein the cells are neutrophils or monocytes.

3. A method of claim 1, wherein the amount effective to inhibit cell stimulation comprises an amount from about 100 ng to about 100 mg.

4. A method of claim 3, wherein the amount effective to inhibit cell stimulation comprises an amount from about 10 ug to about 10 mg.

5. A method of claim 1, wherein the Bactericidal/Permeability Increasing Protein comprises purified Bactericidal/Permeability Increasing Protein.

6. The use of claim 1, wherein the Bactericidal/Permeability Increasing Protein comprises recombinant Bactericidal/Permeability Increasing Protein or a biologically active polypeptide analog thereof.

7. The use of claim 6, wherein the biologically active polypeptide analog of Bactericidal/Permeability Increasing Protein comprises a polypeptide which has a molecular weight of about 25 kD and corresponds to the N-terminal amino acid sequence of Bactericidal/Permeability Increasing Protein.