Novel Neutrophil Inhibitors

Cross Reference to Related Applications

This application is a continuation-in-part of United States Serial No. 08/151,064, filed November 10 1993, which is a continuation in-part of United States Serial No. 08/060,433 filed May 11, 1993, which is a continuation-in-part of United States Serial No. 07/996,972 filed December 24, 1992, which is a continuation-in-part of United States Serial No. 07/881,721 filed May 11, 1992, the disclosures of which are incorporated herein by reference.

Field of the Invention

This invention relates to factors which interact with CDllb/CD18 ingegrin complex or the I-domain portion of CDllb/CD18 integrin complex and inhibit leukocyte activity. These factors inhibit neutrophil activity, including inhibition of neutrophil activation and adhesion of neutrophils to vascular endothelial cells. These factors also inhibit eosinophil activity, including inhibition of eosinophil adhesion to vascular endothelial cells.

Background of the Invention

Leukocytes are a class of cells comprised of lymphocytes, monocytes and granulocytes. The lymphocytes include within their class, T-cells (as helper T-cells and cytotoxic or suppressor T-cell) , B- cells (as circulating B-cells and plasma cells) , third population or natural killer (NK) cells and antigen- presenting cells. Monocytes include within their class, circulating blood monocytes, Kupffer cells, intraglomerular mesangial cells, alveolar macrophages,

serosal macrophages, microglia, spleen sinus macrophages and lymph node sinus macrophages. Granulocytes include within their class, neutrophils, eosinophils, basophils, mast cells, (as mucosa-associated mast cells and connective tissue mast cells) .

Neutrophils are an essential component of the host defense system against microbial invasion. In response to soluble inflammatory mediators released by cells at the site of injury, neutrophils emigrate into tissue from the bloodstream by crossing the blood vessel wall. At the site of injury, activated neutrophils kill foreign cells by phagocytosis and by the release of cytotoxic compounds, such as oxidants, proteases and cytokines. Despite their importance in fighting infection, neutrophils themselves can promote tissue damage. During an abnormal inflammatory response, neutrophils can cause significant tissue damage by releasing toxic substances at the vascular wall or in uninjured tissue. Alternatively, neutrophils that stick to the capillary wall or clump in venules may produce tissue damage by ischemia. Such abnormal inflammatory responses have been implicated in the pathogenesis of a variety of clinical disorders including adult respiratory distress syndrome (ARDS) ; ischemia-reperfusion injury following myocardial infarction, shock, stroke, and organ transplantation; acute and chronic allograft rejection; vasculitis; sepsis; rheumatoid arthritis; and inflammatory skin diseases (Harlan et al., 1990 Immunol. Rev. 114. 5). Neutrophil adhesion at the site of inflammation is believed to involve at least two discrete cell-cell interactive events. Initially, vascular endotheliu adjacent to inflamed tissue becomes sticl for neutrophils; neutrophils interact with the endothelium via low affinity adhesive mechanisms in a process known as "rolling". In the second adhesive step, rolling

neutrophils bind more tightly to vascular endothelial _* cells and migrate from the blood vessel into the tissue. jjt Neutrophil rolling along affected vascular segments

»» and other initial low affinity contacts between

5 neutrophils and the endothelium are reported to be mediated by a group of monomeric, integral membrane glycoproteins termed selections. All three of the selections so far identified, that is L-selectin (LECAM-l or LAM-1) present on the surface of 10 neutrophils, E-selectin (endothelial leukocyte adhesion molecule-l or ELAM-1) present on endothelial cells and P-selectin (granule membrane protein-140, GMP-140, platelet activation-dependent granule-external membrane protein, PADGEM or CD62) expressed on endothelial cells, 15 have been implicated in neutrophil adhesion "to the vascular endothelium (Jutila et al., 1989 J. Immunol 143. 3318; Watson et al. , 1991 Nature 349. 164; Mulligan et al., J. Clin. Invest. £8., 1396; Gundel et al., 1991 J. Clin. Invest. 88_ _/ 1407; Geng et al., 1990 Nature 343, 20 757; Patel et al., 1991 J. Cell Biol. 112., 749). The counter-receptor for E-selectin is reported to be the sialylated Lewis X antigen (sialyl-Lewis ^x ) that is present on cell-surface glycoproteins (Phillips et al., 1990 Science 250, 1130; Walz et al., 1990 Science 250, 25 1132; Tiemeyer et al., 1991 Proc. Natl. Acad. Sci. (USA) 88, 1138; Lowe et al. , 1990 Cell j52, 475). Receptors for the other selections are also thought to be carbohydrate in nature but remain to be elucidated. _, The more stable secondary contacts between 30 neutrophils and endothelial cells are reported to be „ mediated by a class of cell adhesion molecules known as integrins. Integrins comprise a broad range of evolutionarily conserved heterodimeric tra smembrane glycoprotein complexes that are present on virtually all 35 cell types. Members of the leukocyte-specific CD18 (β ₂ ) family of integrins, which include CDlla/CD18 (LFA-1)

and CDllb/CD18 (Mac-1, Mo-1 or CR3) have been reported to mediate neutrophil adhesion to the endothelium (See

Larson and Springer, 1990 Immunol Rev. 114. 181) .

Endothelial cell counter-receptors for these integrins are the intercellular cell adhesion molecules ICAM-1 and ICAM-2 for CDlla/CD18 and ICAM-1 for cmi _b/CD iβ. respectively (Rothlein et al. , 1986 J. Immunol. 137. 1270; Staunton et al., 1988 Cell .52., 925; Staunton et al., 1989 Nature 339, 61) . The ICAMs are monomeric transmembrane proteins that are members of the immunoglobulin superfamily.

The CDllb/CD18 integrin is expressed on a variety of leukocytes, including monocytes, macrophages, granulocytes, large granular lymphocytes (NK cells) , and immature and CD5" ^" B cells (Kishimoto, T.K. , Larson, R.S., Corbi, A.L. , Dustin, M.L. , Staunton, D. ^' E. , and Spriger, T.A. (1989) Adv. in Immunol. 46,149-182). CDllb/CD18 has been implicated in a variety of leukocyte functions including adhesion of neutrophils to endothelial cells (Prieto, J., Beatty, P.G. , Clark, E.A., and Patarroyo, M. (1988) Immunology 63, 631-637; Wallis, W.J., Hickstein, D.D. , Schwartz, B.R. , June, C.H., Ochs, H.D. , Beatty, P.G., Klebanoff, S.J., and Harlan, J.M. (1986) Blood 67, 1007-1013; Smith, C.W. , Marlin, S.D., Rothlein, R. , Toman, C, and Anderson, D.C. (1989) J. Clin. Invest. 83, 2008-2017) and release of hydrogen peroxide from neutrophils (Shappell, S.B., Toman, C. , Anderson, D.C, Taylor, A.A. , Entman, M.L. and Smith, C.W. (1990) J. Immunol. 144, 2702-2711; Von Asmuth, E.J.U. , Van Der Linden, C.J. , Leeuwenberg, J.F.M., and Buurman, W.A. (1991) J. Immunol. 147,3869-

3875) . This integrin may play a roll in neutrophil and monocye phagocytosis of opsonized (ie C3bi-coated) targets (Beller, D.I., Springer, T.A. , and Schreiber, R.D. (1982) J.Exp. Med. 156,1000-1009). It has also been reported that CDllb/CD18 contributes to elevated natural killer activity against C3bi-coated target cells

(Ramos, O.F., Kai, C. , Yefenof, E. , and Klein, E. (1988) J. Immunol. 140,1239-1243).

The activation of endothelial cells and neutrophils is believed to represent an important component of neutrophil-mediated inflammation. Factors that induce cell activation are termed agonists. Endothelial cell agonists, which are believed to include small regulatory proteins such as tumor necrosis factor (TNFα) and interleukin-lα (IL-lα) , are released by cells at the site of injury. Activation of endothelial cells has been reported to result in the increased surface expression of ICAM-1 (Staunton et al., 1988 Cell 52. 925) and ELAM-1 (Bevilacgua et al., 1987 Proc. Natl. Acad. Sci. (USA) .84., 9238) . Raised levels of expression of these adhesive molecules on the surface of activated endothelial cells is believed to lead to the observed increased adhesivity of neutrophils for the vascular endothelium near sites of injury.

Activation of the neutrophil results in profound changes to its physiological state, including shape change, ability to phagocytose foreign bodies and release of cytotoxic substances from intracellular granules. Moreover, activation is believed to greatly increase the affinity of adhesive contacts between neutrophils and the vascular endothelium, perhaps through a conformational change in the CDllb/CD18 integrin complex on the neutrophil surface (Vedder and Harlan, 1988 J. Clin. Invest. 1, 676; Buyon et al., 1988 J. Immunol. 140. 3156) . Factors that have been reported to induce neutrophil activation include IL-lα,

GM-CSF, G-CSF, MIP-1, IL-8 (IL-8 = interleukin-8, GM-CSF = granulocyte/monocyte-colony stimulating factor, G-CSF = granulocyte-colony stimulating factor) , ^TNFα, the complement fragment C5a, the microbe-derived peptide formyl-Met-Leu-Phe and the lipid-like molecules leukotriene B4 (LTB ₄ ) and platelet activating factor

(Fuortes and Nathan, 1992, in Molecular Basis of Oxidative Damage by Leukocytes Eds Jesaitiε, A.J. and Dratz, E.A. (CRC Press) pp. 81-90). In addition, phorbol esters (e.g., phorbol 12-myristate 13-acetate; PMA) have been proposed as a potent class of synthetic lipid-like neutrophil agonists. With the exception of PMA, these agonists are believed to activate neutrophils by binding receptors on their surface. Receptors that are occupied by agonist molecules are believed to initiate within the neutrophil a cascade of events that ultimately will result in the physiological changes that accompany neutrophil activation. This process is known as signal transduction. The lipid-like PMA is proposed to affect neutrophil activation by passing through the plasma membrane at the cell surface and directly interacting with intracellular components (i.e., protein kinase) of the signal transduction machinery.

There exist two general classes of compounds that have been reported to down regulate the function of neutrophils, and these compounds have been shown to mitigate inflammation. One group of anti-inflammatory compounds has been proposed to function as inhibitors of neutrophil activation, and presumably adhesion, by acting on components of the signal transduction machinery. A second class of anti-inflammatory compounds has been proposed to block neutrophil infiltration into inflammatory foci by acting as direct inhibitors of the adhesive receptors that mediate contact between neutrophils and the vascular endothelium.

Many of the anti-inflammatory compounds currently used as therapeutics, including prostaglandins, catecholamines, and a group of agents known as non-steroidal anti-inflammatory drugs (NSAIDs) , are believed to fall into the first category (Showell and Williams, 1989, in Immunopharmacology, eds. Gilman, S.

C. and Rogers, T. J. [Telford Press, NJ] pp 23-63) . For example, the enhanced adhesiveness observed for TNFα-activated neutrophils has been reported as associated with decreased levels of a mediator of signal transduction, cyclic AMP (cAMP) . (See Nathan and

Sanchez, 1990 JCB ill, 2171). Exposure of neutrophils to prostaglandins and catecholamines has been correlated with elevated levels of intracellular cyclic AMP (Showell and Williams, 1989) . While signal transduction inhibitors have been used extensively as anti-inflammatory therapeutic agents, they have been shown to have several disadvantages including poor efficacy in acute inflammatory conditions, lack of specificity and undesirable side-effects such as gastric or intestinal ulceration, disturbances in platelet and central nervous system function and changes in renal function (Insel, 1990 in The Pharmacological Basis of Therapeutics. eds. Gilman, A. G., Rail, T. W. , Nies, A. S., and Taylor, P. [Pergamon, NY], 8th Ed., pp. 638-681) .

Glucocorticoids have long been recognized for their anti-inflammatory properties. Steroid induced inhibition of neutrophils has been reported for several neutrophil functions, including adherence (Clark et al., 1979 Blood 53., 633-641; MacGregor, 1977 Ann. Intern. Med. 8.6, 35-39) . The mechanisms by which glucocorticoids modulate neutrophil function are not well understood, but they are generally believed to involve the amplification or suppression of new proteins in treated neutrophils that play a key role in the inflammatory process (Knudsen et al., 1987 J. Immunol. 139, 4129) . In particular, a group of proteins known as lipocortinε, whose expression is induced^in neutrophils by glucocorticoids, has been associated with anti-inflammatory properties (Flower, 1989 Br. J. Pharmacol. 9_4, 987-1015). Lipocortins may exert

anti-neutrophil effects by interacting with sites on the neutrophil surface (Camussi et al., 1990 J. Exp. Med. 171. 913-927) , but there is no evidence to suggest that the lipocortins act by directly blocking adhesive proteins on the neutrophil. Apart from their beneficial anti-inflammatory properties, glucocorticoids have been associated with significant side-effects. These include suppression of pituitary-adrenal function, fluid and electrolyte disturbances, hypertension, hyperglycemia, glycosuria, susceptibility to infection, ulcers, osteoporosis, myopathy, arrest of growth and behavioral disturbances (Insel, 1990) .

A second class of anti-inflammatory compounds which are reported as direct inhibitors of neutrophil adhesion to the vascular endothelium are monoclonal antibodies. Monoclonal antibodies that recognize and block ligand-binding functions of some of these adhesive molecules have been reported to act as j-n vivo inhibitors of neutrophil-mediated inflammation. In particular, monoclonal antibodies to the CD18 subunit of the CD18 integrin complexes (i.e., CDlla/CD18, CDllb/CD18 and CDllc/CDlδ) on the surface of neutrophils have been reported to prevent a variety of neutrophil-mediated tissue injury in animal models, including pulmonary edema induced by reperfusion (Horgan et al, 1990 Am. J. Physiol. 259. L315-L319) , organ injury induced by hemorrhagic shock (Mileski et al, 1990 Surgery 108, 206-212) , yocardial damage following ischemia/reperfusion (Winguist et al, 1990 Circulation III-701) , edema and tissue damage following ischemia/reperfusion of the ear (Vedder et al, 1990 Proc. Natl. Acad. Sci.(USA) 82, 2643-2646), brain edema and death produced by bacterial meningitis (Tuomanen et al, 1989 J. Exp. Med. 170. 959-968), vascular injury and death in endotoxic shock (Thomas et al, 1991 FASEB J. 5 ,

A509) and indomethacin-induced gastric injury (Wallace et al, 1991 Gastroenterology 100. 878-883).

Monoclonal antibodies directed to the CDllb subunit have been reported by Todd, R.F. et al., U.S. Patent No. 4,840,793 (June 20, 1989), Todd, R.F. et al., U.S.

Patent No. 4,935,234 (June 19, 1990), Schlossman, S.F. et al., U.S. Patent No. 5,019,648 (May 28, 1991) and Rusche, J.R. et al., International Application No. WO 92/11870 (July 23, 1992). Monclonal antibodies directed to CD18 subunit have been reported by Arfors, K.E., U.S. Patent No. 4,797,277 (January 10, 1989), Wright, S.D. et al., European Patent Application No. 346,078 (December 13, 1989), Law, M. et al., European Patent Application No. 438,312 (July 24, 1991), Law, M. et al., European Patent Application No. 440,351 (August 7, 1991), Wright, S.D. et al., U.S. Patent No. 5,147,637 (September 15, 1992) and Wegner, CD. et al., European Patent Application No. 507,187 (October 7, 1992).

Antibodies to other adhesive molecules have also been reported to have anti-inflammatory properties. Monoclonal antibodies that recognize the counter-receptor of CDlla/CD18 and CDllb/CD18, ICAM-1 have been reported to prolong cardiac allograft survival (Flavin et al, 1991 Transplant. Proc. 2_2, 533-534) and prevent chemically induced lung inflammation (Barton et al, 1989 J. Immunol. 143. 1278-1282) . Furthermore, anti-selectin monoclonal antibodies have also been reported as active in animal models of neutrophil-mediated inflammation. Monoclonal antibodies to L-selectin have been reported to prevent neutrophil emigration into inflamed skin (Lewinshon et al., 1987 J. Immunol. 138, 4313) and inflamed ascites (Jutila et al., 1989 J. Immunol. 143, 3318; Watson et a_k.' , 1991 Nature 349, 164) . Reports have also described inhibition of neutrophil influx into inflamed lung tissue by anti

E-selectin monoclonal antibodies (Mulligan et al. , 1991

J. Clin. Invest. 8ji. 1396; Gundel et al., 1991 J. Clin. Invest. , 1407) . While monoclonal antibodies to adhesive proteins have demonstrated the feasibility of using neutrophil adhesion inhibitors as anti-inflammatory agents, their utility as therapeutics requires further evaluation.

Soluble adhesive receptors obtained by genetic engineering have been proposed as anti-inflammatory compounds. Soluble receptors, in which the transmembrane and intracellular domains have been deleted by recombinant DNA technology, have been tested as inhibitors of neutrophil adhesion to endothelial cells. The functional use of recombinant soluble adhesive molecules has been reported using CDllb/CD18 (Dana et al., 1991 Proc. Natl. Acad. Sci. (USA) _^ 88 _. , 3106-3110) and L-selectin (Watson et al., 1991).

Recently, a new class of anti-leukocyte compounds collectively termed "leumedins" has been reported. These compounds have been reported to block the recruitment in vivo of T lymphocytes and neutrophils into inflammatory lesions. The mechanism of action of the leumedins is unclear, but there is evidence that they do not function by blocking neutrophil activation (Burch et al., 1991 Proc. Natl. Acad. Sci. (USA) 88, 355) . It remains to be determined if leumedins block neutrophil infiltration by direct interference with adhesive molecules.

It has been suggested that parasites survive in their host by modulating host immunity and inflammatory response though the mechanisms by which this occurs remains unclear (Leid, W.S., 1987, Veterinary Parasitology, J25: 147) . In this regard, parasite-induced immunosuppression in roςlέnt models has been proposed (Soulsby et al., 1987, Immunol Lett. I6 _r 315-320) . The various aspects of the modulation of host immunity by helminth parasites to evade immunological

attack has recently been reviewed. See Maizels et al. (1993), Nature, 365:797-805.

Various parasites have been reported to have an affect on neutrophils of their host. For example, a protein isolated from the cestode, Taenia taeniaeformis, has been reported to inhibit chemotaxis and chemokinesis of eguine neutrophils, as well as inhibit neutrophil aggregation (C Suquet et al., 1984, Int'l J. Parasitol., .14 _. : 165; Leid, R.W. et al., 1987, Parasite Immunology, 9_: 195; and Leid, R.W. et al., 1987, Int'l J. Parasitol., .17.: 1349). Peritoneal neutrophils from mice infected with the cestode, Echinococcus multiocularis, have been reported to lose their ability to migrate toward parasite antigens and nonspecific chemoattractants with increasing time of infection

(Alkarmi, T. et al., Exptl. Parasitol., 1989, j5£: 16). The nematode, Trichinella spiralis. has been reported to either excrete and/or secrete factors which inhibit chemotaxis and p-nitroblue tetrazolium reduction (i.e., release of oxidative metabolites) but enhance chemokinesis of human neutrophils (Bruschi, F. et al., 1989, Wiadomosci Parazytologiczne, .3j5: 391). The sera of humans infected with the nematode, Trichinella spiralis, has been reported to inhibit leukocyte chemotaxis and phagocytosis (Bruschi, F. et al., 1990, J. Parasitol., 7_6: 577). The saliva of the tick, Ixodes dammini, has been reported to inhibit neutrophil function (Ribeiro et al, 1990, Exp. Parasitol., 70, 382) . A protein secreted by the cestode, Echinococcus granulosus. has been reported to inhibit human neutrophil chemotaxis (Shepard, J.C et al., 1991, Mol. Biochem. Parasitol., : 81).

Another component of the host defence mechanism against invading pathogens are eosinophils. Functionally, eosinophils are similar to neutrophils in that both cell types have the ability to phagocytose and

to release compounds that are either directly or indirectly toxic to pathogenic organisms. Eosinophils are distinguished from neutrophils by their morphologic features, constituents, products and associations with specific diseases. Although eosinophils have been reported to be capable of killing bacteria in vitro, this class of leukocyte alone is not believed sufficient to defend against bacterial infections in vivo. Instead, it is thought that eosinophils afford primary defense against large organisms such as helminthic parasites (Butterworth AE, 1984; Adv. Parasitol. 2J3:143-235) . Also, it is widely held that eosinophils can play a major role in certain inflammatory diseases. Specifically, substances released from eosinophils that are known collectively as cationic granule proteins, including major basic protein, eosinophil cationic protein and eosinophil-derived neurotoxin, have been implicated in asthma (Gleich GJ and Adolphson, CR, 1986; Adv. Immunol. 29_'.177-253) , inflammatory bowel disease (Hallren, R, 1989; Am. J. Med. 6:56-64) and atopic dermatitis (Tsuda, S, et al, 1992; J. Dermatol. 19.:208-213) . Moreover, other eosinophil products such as superoxide anions, hydroxyl radicals and singlet oxygen may also be involved in damage to host tissue in inflammatory disease states (Petreccia, DC et al, 1987,

J. Leukoc. Biol. 41:283-288; Kanofsky, JR et al, 1988; J. Biol. Chem. 263:9692-9696) .

An early step in eosinophil-mediated inflammatory disease is believed to be the movement of eosinophils from the vascular compartment to tissue. The first step in this extravasation process is reported to be the adherence of eosinophils to the luminal surface of the vascular endothelium. Although mechanisms of eosinophil-endothelial cell adhesion are not as well defined as those involving adhesion by neutrophils, it is reported that members of the CD11/CD18 family of

integrins on the surface of the eosinophil are involved in eosinophil-endothelial adhesion (Lamas, AM, et al, 1988; J. Immunol. 140:1500; Walsh, GM, et al, 1990; Immunology 7.1:258) , and it is reported that the endothelial cell counter-receptor is likely ICAM-1 (Wegner, CD, et al, 1990; Science 247:456-459) . A second integrin known as VLA-4 (very late antigen-4, a4bl) that is present on eosinophils, lymphocytes and monocytes but not neutrophils, is thought to contribute to eosinophil adherence by binding to the VCAM-1

(vascular cell adhesion molecule-1) that is expressed on the surface of endothelial cells (Dobrina, A, et al, 1991, J. Clin. Invest. 88.:20). IL-1 treatment of the endothelial cell monolayers has been reported to induce an increased adhesiveness for human basophils _t eosinophils and neutrophils but treatment of these endothelial cells with an antibody directed to VACM-l was reported to inhibit both basophil and eosinophil adhesion but not neutrophil adhesion. It has also been reported that monoclonal antibodies against VCAM-1 inhibit lymphocyte and monocyte cell adhesion to stimulated endothelium (Carlos et al. (1990), Blood, 26:965-970; Rice et al., J. Exp. Med. (1990), 171:1369- 1374) but not to neutrophils. Approaches to the treatment of eosinophil-mediated inflammation have been similar to those adopted for neutrophil-mediated disease. For example, potential therapeutics under investigation for eosinophil-mediated inflammation include glucocorticoids (Evans, PM, et al, 1993, J. Allergy Clin. Immunol. .91:643-650). As is the case for other agents that have been reported to modulate neutrophil function, these agents have been found to be sub-optimal in that they are, _* relatively non-specific and toxic. A second approach to anti-eosinophil therapy has been the use of compounds that directly inhibit the adhesion of eosinophils to

vascular endothelium. It has been reported that in animal models of asthma, monoclonal antibody against ICAM-1 blocks eosinophil infiltration into tissues. Wegner et al. (1990), Science, 247:456-459. ICAM-1 and functional derivatives thereof have been proposed as anti-inflammatory agents. Anderson et al., European Patent Application No. 314,863 (April 29, 1988); Wegner et al., International Application No. WO 90/10453 (September 20, 1990). However, there remains a need for potent, highly specific inhibitors of neutrophil and eosinophil function, in particular, adhesion to vascular endothelium, as a treatment for abnormal granulocyte-mediated inflammation. The present invention describes potent and specific inhibitors of neutrophil and eosinophil activity, in particular the adhesion of these granulocytes to vascular endothelial cells, derived from hookworms (such as Ancylostoma caninum) and related species.

Summary of the Invention

Among other factors, the present invention is based on our finding that the Neutrophil Inhibitory Factor of the present invention represents a pioneering step toward the development of a new generation of anti-inflammatory therapeutic products. This discovery will enable therapy for inflammatory disease based entirely on specific inhibition of the inflammatory response. The therapeutic advantages of this novel approach are realized through the specificity of Neutrophil Inhibitory Factor compared to current clinical treatment modalities such as steroids, catecholamines, prostaglandins, and nonsteroidal anti-inflammatory agents. The currently used therapeutic agents demonstrate poor efficacy and multiple adverse reactions due to generalized systemic

effects that non-specifically target numerous biological processes in addition to the inflammatory process. Nonetheless, the existence of this extensive panel of anti-inflammatory agents, although suboptimal, and the total funds expended by the pharmaceutical industry in research in this area point to significant medical needs for effective anti-inflammatory agents and suggests that the novel and highly specific Neutrophil Inhibitory Factors of the present invention have important applications.

As noted in the Background, the inflammatory response may result in clinical syndromes ranging from debilitating arthritis and asthma to life threatening shock. In view of the severity of these disorders, the vast number of individuals afflicted therewith and the lack of suitable therapeutic intervention, the need for a breakthrough therapy represents a long felt need which has not been met. The Neutrophil Inhibitory Factor of the present invention is believed to meet this need by providing the potential for a lifesaving therapy which is currently being sought throughout the international medical and pharmaceutical research communities.

Further, in view of the myriad conditions associated with undesired and/or abnormal inflammatory conditions which appear to be associated with neutrophil activity, there remains a need for potent, highly specific inhibitors of neutrophil function, in particular, adhesion to vascular endothelium, as a treatment for abnormal neutrophil-mediated inflammation. The present invention is believed to fulfill this need by disclosing a potent and specific inhibitor of neutrophil activity, in particular the adhesion of neutrophils to vascular endothelial cell_έ', derived from hookworms (such as Ancylostoma caninum) and related species.

The present invention is directed to a neutrophil inhibitory factor ("Neutrophil Inhibitory Factor" or "NIF") which may be isolated from natural sources or made by recombinant methods. Neutrophil Inhibitory Factor is a protein which is neither an antibody, a member of the integrin or selectin families nor a member of the immunoglobulin superfamily of adhesive proteins and which when isolated from a parasitic worm is a glycoprotein. Recombinant NIF produced by certain expression systems may or may not be glycosylated, or may be glycosylated to a variable degree. However, such NIFs whether glycosylated or not are considered to be within the scope of the present invention.

In one aspect, the present invention provides NIFs which contain, as part of their total amino acid sequence, an amino acid sequence selected from the group consisting of

(a) Arg-Xι-X ₂ -Phe-Leu-X ₃ -X ₄ -His-Asn-Gly-Tyr-Arg-Ser- X ₅ -Leu-Ala-Leu-Gly-His-X ₆ -X ₇ -Ile, wherein X- is Leu or Arg; X ₂ is Gin, Lys or Arg; X ₃ is Ala or Arg; X ₄ is Leu or Met; X ₅ is Lys, Arg, Leu or lie; X ₆ is Val or lie; and X ₇ is Ser, Gly or Asn;

(b) Ala-X ₈ -X ₉ -Ala-Ser-X ₁₀ -Met-Arg-X,.-Leu-Xι ₂ -Tyr- Asp-Cys-χ. ₃ -Ala-Glu-Xi ₄ -Ser-Ala-Tyr-X ₁₅ -Ser-Ala , wherein X ₈ is His or Pro ; X ₉ is Thr, Arg or Ser ; X ₁₀ is Arg or Lys ;

X _n is lie or Tyr; X ₁₂ is Asp, Lys or Glu; X ₁₃ is Asp or Glu ; X ₁₄ is Gly, Lys or Arg; and X ₁₅ is Glu, Met, Thr or Val ;

(c) Ser-X ₁₆ -Phe-Ala-Asn-X ₁₇ -Ala-Trp-Asp-X ₁₈ -Arg- Glu-Lys-X ₁₉ -Gly-Cys-Ala-Val-Val-X ₂₀ -Cys , wherein X ₁₆ is Asn or Asp; X ₁₇ is Val or Leu; X- ₈ is Ala or Thr; X ₁₉ is Leu , Val or Phe ; and X ₂₀ is Thr, Lys or Asn;

(d) His-Val-Val-Cys-His-X _2l -X ₂₂ -Pro^Lys , wherein X ₂₁ is Tyr or lie ; X ₂₂ is Gly or no residue; (e) Ile-Tyr-Xy-X^-Gly-X^-Pro-Cys-X^-X^-Cys-X^-

X ₂₉ -Tyr , wherein X ₂₃ is Thr , Ser , Lys or Glu; X _u is Thr ,

Val or lie; X^ is Val, Lys or Thr; X ₂₆ is Arg, Ser or Asp; X ₂₇ is Asn, Gly, Asp or Arg; X ₂₈ is Asn, Ser or Thr; and X ₂₉ is Gly, Glu or Asp; and

(f) Cys-X ₃₀ -X ₃₁ -Asp-X ₃₂ -Gly-Val-Cys-X ₃₃ -Ile, wherein X ₃₀ is His, lie or Asn; X ₃ ι is Ala, Pro or Asp; X ₃₂ is Glu, Val, Asp or lie; X ₃₃ is lie, Val or Phe. Such NIFs exhibit neutrophil inhibitory activity.

In another aspect, the present invention provides mutant NIFs wherein certain asparagine residues are replaced with glutamine residues which is believed to result in the reduced glycosylation of these NIFs. The mutant NIFs contain, as part of their total amino acid sequence, an amino acid sequence selected from the group consisting of peptides (a) to (f) hereinabove. Such NIFs exhibit neutrophil inhibitory activity. -

In another aspect, the present invention provides NIFs which contain, as part of their total amino acid seguence, an amino acid sequence encoded by a nucleic acid sequence which is sufficiently complementary to hybridize to certain nucleic acid probes. Such NIFs exhibit neutrophil inhibitory activity. The present invention includes within its scope these nucleic acid probes.

In another aspect, the present invention provides nucleic acid molecules encoding for a NIF and which are isolated as described herein. Such isolated nucleic acid molecules include expression vectors containing a nucleic acid sequence encoding a NIF. The present invention also includes the host cells transformed by such expression vectors.

In another aspect, the present invention provides methods for making biologically active NIFs, wherein such NIFs are expressed and, optionally^secreted. The present invention also includes the NIFs made by these methods.

In another aspect, the present invention provides methods of making NIFs comprising preparing a cDNA library from a source suspected of having a NIF and hybridizing certain oligonucleotide probes of the present invention to the nucleic acid molecules from the source. Such NIFs exhibit neutrophil inhibitory activity. The present invention also includes the NIFs made by these methods.

In another aspect, the present invention provides methods of detecting in a sample the presence of a nucleic acid molecule encoding a NIF, which methods comprise the combining the sample thought to contain such nucleic acid molecule with a probe of the present invention and detecting the presence of hybridized probe.

In another aspect, the present invention provides monoclonal antibodies which bind to NIF. The present invention also includes the hybridoma cell lines which make such antibodies, a method of purifying NIF using such monoclonal antibodies and method of detecting in a sample the presence of NIF using such antibodies.

In another aspect, the present invention provides a method for detecting in a sample NIF mimics which compete with NIFs for binding to the CDllb/CD18 receptor, which method comprises contacting a sample with the CDllb/CD18 receptor. Also provided is a method for detecting in a sample NIF mimics which compete with NIFs for binding to the I-domain portion of the CDllb/CD18 receptor, which method comprises contacting a sample with a recombinant peptide comprising the I- domain of CDllb/CD18 receptor. Such NIF mimics exhibit neutrophil inhibitory activity. The present invention also includes the NIF mimics detected b# ^' these methods.

In another aspect, the present invention provides a method for detecting in a sample a NIF antagonist which

prevents NIF binding to the CDllb/CD18 receptor, which method comprises contacting such sample with the CDllb/CD18 receptor. Also provided is a method for detecting in a sample NIF antagonist which compete with NIFs for binding to the I-domain portion of the

CDllb/CD18 receptor, which method comprises contacting a sample with a recombinant peptide comprising the I- domain of CDllb/CD18 receptor. Such NIF antagonists do not exhibit neutrophil inhibitory activity themselves. The present invention also includes the NIF antagonists detected by these methods.

In another aspect, the present invention provides methods of using NIF to treat inflammatory conditions, especially to prevent or decrease inflammatory responses, which methods comprise administering to a mammal a therapeutically effective amount of NIF. Other features and advantages of the present invention will be apparent from the following descriptions of the preferred embodiments and from the claims.

Definitions

In accordance with the present invention and as used herein, the following terms are defined with the following meanings, unless explicitly stated otherwise. The term "amino acid" refers to the natural L-amino acids. The natural amino acids shall be referred to by their names or may be abbreviated as shown below: L-amino acid Abbreviation Three letter One Letter alanine Ala A arginine Arg R asparagine Asn N aspartic acid Asp D cysteine Cys C glutamine Gin Q

glutamic acid Glu E glycine Gly G histidine His H isoleucine lie I leucine Leu L lysine Lys K methionine Met M phenylalanine Phe F proline Pro P serine Ser S threonine Thr T tryptophan Trp W tyrosine Tyr Y valine Val V

The term "amino acid residue" refers to

-NH-CH(R)-CO-, wherein R is the side chain group distinguishing each amino acid. For cyclic amino acids, the residue is (CH ₂ ) _X

<

N

I

wherein x is 1, 2 or 3 representing the azetidine- carboxylic acid, proline or pipecolic acid residues, respectively.

The term "isoform" refers to a family of related proteins from a single organism having homologous sequences of amino acid residues interspersed with variable sequences.

The term "nucleic acid" refers to polymers of either deoxynucleic acids or ribonucleic*acids, either single-stranded or double-stranded.

The term "isolated nucleic acid" refers to nucleic acids which are isolated by biochemical or molecular

biology techniques such as centrifugation, chromatography, electrophoresis, hybridization and the like.

The term "NIF mimic" refers to a small molecule, peptide, peptide analog or protein, which competes with NIF for binding to the CDllb/CD18 receptor or the I- domain portion of the CDllb/CD18 recptor. A NIF mimic is also characterized as having neutrophil inhibitory activity, eosinophil inhibitory activity or both such activities.

The term "NIF antagonist" refers to a small molecule, peptide, peptide analog or protein, which prevents the binding of NIF to the CDllb/CD18 receptor or the I-domain portion of this receptor, and. does not possess any significant neutrophil inhibitory activity. A NIF antagonist prevents binding of NIF to the CDllb/ CD18 receptor or the I-domain portion of the CDllb/CD18 receptor, by binding to a site on NIF which is required for binding to the receptor in effect sterically hindering binding, or alternatively, by binding to a site on NIF which results in a conformational change to the site needed for such binding which change substantially weakens or abolishes binding.

Brief Description of the Drawings Figure 1 depicts a chromatogram of hookworm lysate obtained as described in the Example 2(A) run on the Example 2(B) Concanavalin A Sepharose column.

Figure 2 depicts a chromatogram of Concanavalin A-purified hookworm lysate run on the Example 2(C) Superdex 200 column.

Figure 3 depicts a chromatogram of<*the Concanavalin A Sepharose/Superdex purified hookworm lysate run on the Example 2(D) ceramic hydroxyapatite column.

Figure 4 depicts a chromatogram from reverse phase HPLC of hookworm lysate isolated by Concanavalin A Sepharose, Superdex 200 and hydroxyapatite chromatography as described in Example 1(E). Figure 5 depicts a gel pattern run using SDS-gel electrophoresis of the HPLC isolate and certain molecular weight standards.

Figure 6 depicts laser-desorption time-of-flight mass spectrometry of the purified Neutrophil Inhibitory Factor of the present invention.

Figure 7 depicts the amino acid sequence of proteolytic fragments prepared from Neutrophil Inhibitory Factor isolated from canine hookworms.

Figure 8 depicts the nucleotide sequence of the coding region of Neutrophil Inhibitory Factor CDNA (clone 1FL) and its predicted amino acid sequences.

Figure 9 depicts the alignment of the predicted amino acid sequences of several Neutrophil Inhibitory Factor isoform clones. Figure 10 depicts the anti-inflammatory effect of varied doses of Neutrophil Inhibitory Factor isolated from canine hookworms administered intraperitoneally in an animal model of inflammation.

Figure 11 depicts the anti-inflammatory effect of Neutrophil Inhibitory Factor isolated from canine hookworms administered either intraperitoneally or intravenously in an animal model of inflammation.

Figure 12 depicts the anti-inflammatory effect of recombinant Neutrophil Inhibitory Factor produced in Pichia pastoris administered in vivo in an animal model of inflammation.

Figure 13 depicts the effect of recombinant NIF on the inhibition of eosinophil adherence *fc ^' o TNF-stimulated HUVEC monolayers. Data points are means of triple determinations of one experiment.

Figure 14 depicts genetic map of the expression vector Pma5-NIl/3. The vector contains the following elements: (i) a ColEl type origin of replication (ORI); (ii) the intercistronic region of filamentous phage fl including the origin of replication (fl ORI) ; (iii) the beta-lactamase gene which confers resistance to ampicillin (bla) ; (iv) the chloramphenicol acetyl transferase gene which contains an amber translational stop codon as the result of a single nucleotide substitution (cat-am) ; (v) the phage lambda P _R promoter; (vi) a small leader cistron; (vii) the methionyl-NIF encoding region and (viii) two tandemly arranged copies of the central transcription terminator of phage fd (fdT) . The sequence of the Met-NIF expression cassette (blown-up region) is shown in Figure 15. The pma/c family of vectors have been described. See Stanssens et al., Nucl. Acids Res. 17, 4441-4454, 1989.

Figure 15 depicts the nucleotide base sequence of the two-cistron Met-NIF expression cassette of Pma5-NIl/3. The encoded methionyl-NIF and leader peptide are shown in the one-letter code. The following features are indicated: the -35 and -10 regions of the phage lambda P _R promoter, the transcription initiation point, the Shine-Dalgarno elements preceding the leader cistron (SD _cro ) and the Met-NIF gene (SD) and some restriction sites. The vector was obtained by ligation of the PCR amplified NIF-1FL coding region to the recipient vector which was cleaved by Ncol, treated with the Klenow fragment of DNA polymerase I and subsequently digested with Hindlll (both ligation points are indicated) . The construction scheme fuses the 5'-end of the NIF coding region to an ATG initiator codon.

Figure 16 depicts a comparison of the nucleotide and amino acid sequences of NIF proteins from hookworms. The NIF-lFL nucleotide sequence is numbered; this numbering is also used to refer to positions in other

genes. PCR-NIF7 and PCR-NIF20 were recovered by PCR-technology: the (regions matching with the) PCR-primers are italicized. AcaNIF7 and NIF-1FL9 differ at only one position (G to E replacement; nucleotide-substitution located at position 647) . The one remaining uncertainty in these sequences is at position 660 in PCR-NIF20. No poly(A+) tail was found in NIF-1FL sequence. Underlined sequences in the 3'-untranslated region (UAUAAA and AGUAAA) may serve as polyadenylation signals. Only the NIF-1FL, NIF-1FL9 and AcaNIF24 cDNAs contain an entire secretion signal. The potential N-glycosylation sites (N-X-T/S) are underlined.

Figure 17 depicts a comparison of the potency of recombinant NIF-1FL with that of the recombinant proteins AcaNIF6 and AcaNIF24. The three purified proteins were tested in both the hydrogen peroxide release assay of Example 1(E) (panel A) and the neutrophil-plastic adhesion assay of Example 1(C) (panel B) .

Figure 18 depicts a comparison of recombinant NIF-1FL with recombinant AcaNIF4 in the competition binding assay of Example 1(F). Samples containing a fixed amount of biotinylated NIF-1FL and varying amounts of either AcaNIF4 or NIF-1FL, were assayed on immobilized LM2/Mac-1 complex. The amount of bound biotinylated NIF-lFL was detected with ExtrAvidin conjugated with alkaline phosphatase.

Figure 19 depicts the nucleotide and amino acid sequence of a NIF protein from A. ceylanicum (AceNIF3) .

The underlined sequence (GAATTCCG) derives from the EcoRI linker that was added onto the CDNA. The sequence which may function as polyadenylation signal (AAUAAA) is also indicated. The encoded protein contains 10 potential N-glycosylation sites (N-X-T/S) .

Figure 20 depicts the binding of phages displaying a NIF protein from A. ceylanicum (AceNIF3) to Mac-l. Wells coated with LM2/Mac-l complex were first incubated with varying amounts of Pichia-produced RNIF1 (or buffer in the control experiment) ; after 30 minutes, 10 ¹⁰ virions displaying either NIF-IFL or AceNIF3 were added and the incubation continued for another 90 minutes. Retained phages were detected with rabbit anti-phage serum and goat anti-rabbit alkaline phosphatase conjugate. The various components were incubated in PBS containing 1 Mm MgCl ₂ , 1 Mm CaCl ₂ and 0.1% skim-milk (phage samples contained 1% skim-milk) . Unbound material was removed by washing with PBS containing 1 Mm CaCl ₂ , 1 Mm MgCl ₂ , 0.1% Tween 20 and 0.02% thimerosal. The relative amount of bound phages was calculated by taking the OD _405nm reading obtained in the control experiment as 100%.

Figure 21 depicts the synthesis of functional NIF-IFL by PAN-NIF-1FL in either a phage-attached or 'soluble' form. The PAN-NIF-1FL vector contains the following elements:

1) a NIF-IFL containing gene fusion which is placed under the transcriptional control of the Plac promoter. The gene fusion consists of the pelB secretion signal, the NIF-IFL coding region, and the filamentous phage M13 genelll (gill) . Between the NIF-IFL and the GUI seguences a TAG amber translational stop codon is present. The Ncol and Notl restriction sites used for the cloning of the NIF-IFL coding region are indicated. 2) the bla gene which confers resistance to lOOμg/ml ampicillin or carbenicillin (bla/Ap ^R ) .

3) the intergenic region, including the origin of replication, of filamentous phage M13 (M13-IR/0RI) .

4) a ColEl-type plasmid borne origin of replication (ColEl-ORI) .

In su ⁺ bacteria such as TGI, the PAN-NIF-1FL phagemid-vector codes for a NIF-lFL-pglll fusion protein (pgiii; product of GUI) which becomes incorporated into filamentous virions upon infection of TG1[PAN-NIF-1FL] cells with M13-VCS 'helper'-phage (Statagene) . In su' strain WK6, PAN-NIF-1FL directs the synthesis of 'free' (not phage-attached) RNIF1 which binds to the anti-NIF MAb 3D2 and to Mac-l. TGI: Δ(lac-proAB) , hsdΔ5 (r _κ Tn _κ ") , thi, supE / F'[traD36, lacl ^q , lacZΔM15, proA ⁺ B ⁺ ].

WK6: Δ(lac-proAB) , galE, strA / F' [lacl ^q , lacZΔM15, proA ⁺ B ⁺ ] .

Figure 22 depicts ELISAs with NIF-lFL-displaying phage. NIF-displaying phages (or non-displaying control phages, e.g. M13-VCS; in each experiment about 5xl0 ⁹ phage particles were added) were incubated in microtiter wells coated with Mac-l (immunopurified on LM2-Sepharose) , LM2/Mac-1, LM2 or the non-neutralizing anti-NIF Mab 3D2. In some experiments non-phage-attached 'soluble' NIF (sNIF) was allowed to react with the immobilized material (1 μM) 30 minutes prior to addition of the phages. Bound phages were detected with a rabbit anti-M13 serum and an alkaline phosphatase conjugated goat anti-rabbit serum. Figure 23 depicts a comparison of the potency of

AcaNIFl with that of the mutants, AcaNIFl/Δh,Gll-7 and AcaNIFl/ΔGll-5, in a neutrophil-plastic adhesion assay of Example 1(C) .

Figure 24 depicts binding of biotinylated rNIF to lymphocytes, monocytes and granulocytes as described in

Example 30.

Figure 25 depicts direct concordance of NIF binding and CDllb/CD18 expression as determined*by dual fluorescence analysis flow cytometry of peripheral lymphocyte populations as described in Example 31.

Detailed Description of the Invention 1. Neutrophil Inhibitory Factor.

The present invention in its various aspects is directed to Neutrophil Inhibitory Factor ("NIF"), a protein that inhibits neutrophil activity and which is not an antibody, an integrin, a selectin or a member of the immunoglobulin superfamily of adhesive proteins and which, when isolated from a parasitic worm, is a glycoprotein. Recombinant NIFs produced by certain expression systems are not glycosylated. Such non-glycosylated NIFs are considered to be within the scope of the invention.

The inhibition of neutrophil activity by the NIFs of the present invention includes but is not limited to inhibition of one or more of the following aαtivities by neutrophils: release of hydrogen peroxide, release of superoxide anion, release of myeloperoxidase, release of elastase, homotypic neutrophil aggregation, adhesion to plastic surfaces, adhesion to vascular endothelial cells, chemotaxis, transmigration across a monolayer of endothelial cells and phagocytosis. Preferred assays include those where inhibition of neutrophil activity is demonstrated by an in vitro assay which determines adhesion of neutrophils to vascular endothelial cells, release of hydrogen peroxide from neutrophils, homotypic neutrophil aggregation or adhesion of neutrophils to plastic surfaces. Preferred NIFs would have an IC ₅₀ of about 500 Nm or less, more preferably less than 100 N , as measured by one of these neutrophil activity assays. An IC ₅₀ is that concentration of a NIF giving 50% inhibition of the measured activity (see Example 1) .

The NIFs of the present invention are further characterized as also having the ability* ^' to bind to the CDllb/CD18 receptor (see Example 14) . Preferred assays for determining the binding of NIF to CDllb/CD18 is described in Example 1(F).

The NIFs of the present invention are further characterized as also having the ability to bind to the I-domain portion of the CDllb/CD18 receptor (see Example 32) . A preferred assay for determining the binding of the NIF to the I-domain portion is described in Example 32.

The NIFs of the present invention may be further characterized as having eosinophil inhibitory activity. A preferred assay for determining eosinophil inhibitory activity is the inhibition of eosinophil activity demonstrated by an in vitro assay which determines adhesion of neutrophils to vascular endothelial cells as described in Example 29. A preferred NIF would have an IC _j0 of about 500 Nm or less, more preferably less than 100 Nm, as measured by this eosinophil activity assay.

(a) Enriched Compositions.

In another aspect, the present invention is directed to compositions enriched for NIF, comprising NIF and which are a isolated by chromatographic or molecular biology methods, or a combination of both methods, from a parasitic worm, preferably a nematode.

Suitable parasitic worms include those selected from species of the phyla Platyhelminthes, Nematoda, Nematomorpha or Acanthocephala. An especially preferred source is endoparasitic hookworm species, such as those found to infect canines. It is believed that certain isoforms of NIF are produced by canine hookworm Ancylostoma species such as Ancylostoma caninum. Another suitable source is the endoparasitic worm species Toxocara canis. Substantially similar compounds may be isolated from other nematode species, as well as from other endoparasites of other phyla* Preferred sources for NIF include parasites, including parasitic worms, particularly endoparasitic nematodes and especially hookworm species, including Ancylostoma

braziliense. Ancylostoma caninum. Ancylostoma ceylanicum. Ancylostoma duodenale, Ancylostoma iaponica. Ancylostoma malavanum. Ancylostoma tubaeforme. Bunostomum phlebotomum. Cyclodontostomum purvisi _r Necator americanus, Necator arqentinus. Necator suillus, and Uncinaria stenocephala.

The enriched compositions may be enriched for NIF in one aspect by chromatographic methods, which methods may include chromatography on Concanavalin A Sepharose ^® , hydroxyapatite or an anion exchange column, gel filtration chromatography preferably using Superdex ^® 200, C4 reverse phase HPLC, isoelectric focusing or a combination of those methods or equivalent methods used for separating proteins or proteinaceous factors. For example, in place of Concanavalin A, other immobilized lectins may be used. In place of Superdex ^® 200, other acrylamide- or agarose-based gel filtration media which fractionate proteins in the appropriate molecular weight range may be used; these include those sold under the tradenames, Sephacryl ^® and Superose ^® (Pharmacia) .

According to a preferred embodiment, the enriched composition is comprised of NIF. The NIF therein is a glycoprotein derived from or isolated from a parasitic worm, preferably a nematode, and more preferably a hookworm species, especially a canine hookworm species or, alternatively, a Toxocara species, or a compound, preferably a protein, which is substantially similar to said glycoprotein. It is believed that certain isoforms of said glycoprotein are produced by the canine hookworm Ancylostoma caninum. By substantially similar is meant that the compound exhibits selective neutrophil inhibitory activity similar to that of the glycoprotein, and, preferably has an IC ₅₀ of about 500^Nm or less, more preferably less than 100 Nm, as measured by neutrophil activity assays such as those described herein.

(B) Glycoprotein NIF and Isoforms.

In another aspect, the NIFs of the present invention comprise a purified glycoprotein. A NIF may be determined to be a glycoprotein by evaluating binding to Concanavalin A Sepharose (see Example 2(B)) and by positive testing as a glycoprotein in GlycoTrack™ diagnostic assay for the presence of carbohydrate groups (see Example 7) .

One glycoprotein having neutrophil inhibitory activity which was isolated from canine hookworms has the following characteristics: This glycoprotein is acidic and exhibits an isoelectric point of about 4.5 as determined by isoelectric focusing (see Example 3) . It has an observed molecular weight of about 41,000 daltons (± 3,000) as determined by laser-desorption _. time-of-flight mass spectrometry (see Example 6) . Its behavior when subjected to SDS-polyacrylamide gel electrophoresis indicated that it contained multiple disulfide bonds, since the reduced glycoprotein migrated on the gel at a significantly higher apparent molecular weight (see Example 5) . The glycoprotein was demonstrated to specifically inhibit neutrophil activity and not to act as a general cytotoxin in another cell adhesion assay (see Example 13) . This glycoprotein was demonstrated to inhibit neutrophil adhesion to vascular endothelial cells and homotypic neutrophil aggregation. One such enriched composition (see Example 2(D)) exhibited an IC ₅₀ of about 10 Nm. An IC ₅₀ is that concentration of inhibitor giving 50% inhibition of the measured activity (see Example 1) . This glycoprotein was demonstrated to inhibit peritoneal inflammatory response when administered intraperitoneally or intravenously in an animal model of acute inflammation (see Example 16) . This enriched composition was demonstrated to inhibit hydrogen peroxide release from neutrophils (see Example 1(E)) and neutrophil adhesion/spreading on plastic (see

Example 1(C)). The Example 2(D) preparation had an IC ₅₀ of about 10 Nm. An enriched composition of the neutrophil function inhibitory factor was shown to have no inhibitory effect on platelet aggregation (see Example 13) .

A second glycoprotein having neutrophil inhibitory activity has been isolated from Toxocara canis. This glycoprotein has an observed molecular weight of about 20,000 daltons as determined by molecular sieve chromatography. This glycoprotein was demonstrated to inhibit neutrophil adhesion to vascular endothelial cells and neutrophil adhesion/spreading on plastic.

In another aspect, the present invention is directed to methods for making enriched compositions comprising NIF, wherein such compositions are isolated from natural sources which may include but are not limited to the parasitic worms.

One preferred embodiment comprises isolating these enriched compositions by chromatographic methods, which methods would include chromatography on Concanavalin A- Sepharose ^® , hydroxyapatite or an anion exchange column, gel filtration chromatography preferably using Superdex ^® 200, C4 reverse phase HPLC, isoelectric focusing or a combination of those methods or equivalent methods used for separating proteins or proteinaceous factors. For example, in place of Concanavalin A-, other immobilized lectins may be used. In place of Superdex ^® 200, other acrylamide- or agarose-based gel filtration media which fractionate proteins in the appropriate molecular weight range may be used; these include those sold under the tradenames, Sephacryl ^® and Superose ^® (Pharmacia) . Preferred methods of preparing enriched compositions comprising NIF are provided which comprise subjecting a lysate from a parasitic worm to the following isolation steps (a) chromatography on Concavalin-A Sepharose ^® , and (b) gel filtration on Superdex ^® 200, and (c)

chromatography on ceramic hydroxyapatite. The enriched composition may be further enriched for NIF subjecting it to the further isolation step of reverse phase high performance liquid chromatography (HPLC) using a C4 column. Examples of methods of preparing the enriched compositions according to the present invention are described in Examples 2 to 5.

The enriched compositions comprising NIF isolated by chromatographic methods are at least about 50% pure, that is, they contain at least about 50% NIF.

Preferably, the composition is enriched at least about 200-fold. According to another preferred embodiment, substantially pure Neutrophil Inhibitory Factor is prepared. By "substantially pure" is meant at least about 90 percent pure. More preferably the Neutrophil Inhibitory Factor so prepared is chromatographically pure.

It is believed that a NIF isolated from a particular source may include multiple isoforms. Accordingly such isoforms are considered to be scope of the present invention. Accordingly, in another aspect, the present invention is directed to a NIF which includes an amino acid sequence selected from the group consisting of (a) Arg-X ₁ -X ₂ -Phe-Leu-X ₃ -X ₄ -His-Asn-Gly-Tyr-Arg-Ser-

X ₅ -Leu-Ala-Leu-Gly-His-X ₆ -X ₇ -Ile, wherein X- is Leu or Arg; X ₂ is Gin, Lys or Arg; X ₃ is Ala or Arg; X ₄ is Leu or Met; X ₅ is Lys, Arg, Leu or lie; X ₆ is Val or lie; and X ₇ is Ser, Gly or Asn; (b) Ala-X ₈ -X ₉ -Ala-Ser-X ₁₀ -Met-Arg-X _π -Leu-X ₁₂ -Tyr-

Asp-Cys-X ₁₃ -Ala-Glu-X ₁₄ -Ser-Ala-Tyr-X ₁₅ -Ser-Ala, wherein X ₈ is His or Pro ; X ₉ is Thr, Arg or Ser; X, ₀ is Arg or Lys ; X _n is lie or Tyr ; X _π is Asp , Lys or Glu«f' X ₁₃ is Asp or Glu ; X ₁₄ is Gly , Lys or Arg; and X ₁₅ is Glu, Met, Thr or Val ;

(c) Ser-X ₁₆ -Phe-Ala-Asn-X ₁₇ -Ala-Trp-Asp-X. ₈ -Arg- Glu-Lys-X ₁₉ -Gly-Cys-Ala-Val-Val-X ₂₀ -Cys , wherein X ₁₆ is Asn or Asp; X _π is Val or Leu; X _J8 is Ala or Thr; X ₁₉ is Leu , Val or Phe ; and X ₂₀ is Thr, Lys or Asn; (d) His-Val-Val-Cys-His-X ₂₁ -X ₂₂ -Pro-Lys , wherein X ₂₁ is Tyr or lie ; X ₂₂ is Gly or no residue;

(e) Ile-Tyr-X _y -X _M -Gly-Xy-Pro-Cys-X^-X^-Cys-X _j g- X ₂₉ -Tyr , wherein _M is Thr, Ser, Lys or Glu; _u is Thr , Val or lie ; X^ is Val , Lys or Thr ; X ₂₆ is Arg, Ser or Asp ; X ₂₇ is Asn, Gly, Asp or Arg; X ₂₈ is Asn, Ser or Thr ; and X ₂₉ is Gly, Glu or Asp; and

(f) Cys-X ₃₀ -X ₃₁ -Asp-X ₃₂ -Gly-Val-Cys-X ₃₃ -Ile, wherein X ₃₀ is His , lie or Asn; X is Ala, Pro or Asp; X ₃₂ is Glu, Val , Asp or lie; X ₃₃ is lie, Val or Phe. The NIF may include from 1 to 6 of amino acid sequence (a) to (f) above. Preferably the NIF includes all of amino acid sequences (a) to (f) . Preferably, the listed amino acid sequences appear in the following order in the protein (from amino terminal end to carboxy terminal end) : (a) , (b) , (c) , (d) , (e) , (f) . Additional amino acid residues or peptide sequences may be interspersed between the above sequences or may be located at the amino terminal and/or carboxy terminal end of the protein (for example, see Figures 7 and 8) . The amino acid sequences of some of the NIF isoforms containing one or more of (a) , (b) , (c) , (d) , (e) or (f) are shown in Figures 9 for canine hookworm NIF, in Figure 16 for Ancylostoma caninum NIF and in Figure 19 for Ancylostoma ceylanicum NIF. Preferred are NIFs comprising the amino acid sequence depicted in Figure 8.

According to the present invention, included are NIFs which have one or more of amino acid sequences (a) , (b) , (c) , (d) , (e) and (f) and exhibit neutrophil inhibitory activity in at least one of the in vitro assay. Suitable assays include those assays which determine adhesion of neutrophils to vascular

endothelial cells, release of hydrogen peroxide from neutrophils, homotypic neutrophil aggregation and adhesion of neutrophils to plastic surfaces. Preferred are NIFs which have an IC ₅₀ of about 500 Nm or less, more preferably less than 100 Nm, as measured by one these neutrophil activity assays and which do not substantially inhibit platelet aggregation at such neutrophil inhibitory concentrations. An IC ₅₀ is that concentration of a NIF giving 50% inhibition of the measured activity (see Example 1) .

The NIFs of the present invention are further characterized as also having the ability to bind to the CDllb/CD18 receptor (see Example 14) . Preferred assays for determining the binding of NIF to CDllb/CD18 receptor are described in Example 1(F).

The NIFs of the present invention are further characterized as also having the ability to bind to the I-domain portion of the CDllb/CD18 receptor (See Example 32) . A preferred assay for determining the binding of NIF to the I-domain portion is described in Example 32.

The NIFs of the present invention are also further characterized as having eosinophil inhibitory activity.

A preferred assay for determining eosinophil inhibitory activity is the inhibition of eosinophil activity demonstrated by an in vitro assay which determines adhesion of neutrophils to vascular endothelial cells as described in Example 29. Preferred are NIFs having an IC _S0 of about 500 Nm or less, more preferably less than 100 Nm, as measured by this eosinophil activity assay.

(C) Recombinant NIFs.

In another aspect, the present invention is directed to NIFs made by methods comprising hybridizing the nucleic acid molecules from a source suspected to contain a NIF to certain oligonucleotide primers or CDNA made from such primers. Such NIFs exhibit neutrophil

inhibitory activity. In yet another aspect, the present invention is directed to these methods of making NIFs. In one preferred aspect, the present invention is directed to NIFs comprising an amino acid sequence which is encoded by a nucleic acid sequence which is sufficiently complementary to hybridize to a primer derived from the amino acid sequence of a NIF. Preferred in the PCR cloning method are single stranded DNA primers of 20-100 nucleotides derived from the sequence of NIF from Ancylostoma canium. Preferred are primers having the following characteristics: limited degeneracy; adherence to codon usage preferences of the particular species from which the library is constructed and primers that target sequences which are conserved among the twelve Ancylostoma caninum NIF isoforms. Each PCR reaction utilizes two primers: a 5-primer that corresponds to the sense strand and a 3'-primer that corresponds to the antisense strand of the NIF coding seguence. Especially preferred are the primers 5-CTCGAATTCT(GATC)GC(ATC)AT(ATC) (CT)T(GATC)GG(ATC)TGGGC- 3' and

5'-CTCGAATTCTT(TC)TCTGG(GA) A(GA)CG(GA)TC(GA) A-3' . The nucleotides within enclosing parentheses are redundant in that any one nucleotide may be used at the position enclosed by such parentheses.

The nucleic acid sequence of the DNA primers are preferably derived from the sequence of NIF from Ancylostoma canium. As described above, one example of NIF of the present invention comprises a glycoprotein which has been isolated in substantially pure form.

Using procedures known in the art, one of ordinary skill in the art can use this protein to derive its amino acid sequence. For example, the protein may^be analyzed to determine an N-terminal sequence, or fragments of the protein can be produced by enzymatic or other specific digestion procedures and the sequence of the terminal

amino acids of those fragments determined. Such amino acid sequences, even if only between five and six contiguous amino acids in length, will provide sufficient information to determine potential DNA sequences of a gene encoding this protein.

If two or three such amino acid fragments are sequenced, a plurality of oligonucleotides can be synthesized using reported procedures, and such oligonucleotides can be used to probe a genomic or CDNA library from hookworm (or other source) to isolate the gene or fragments thereof encoding the sequenced protein. Those in the art will recognize that these oligonucleotides can be designed using standard parameters such that the oligonucleotide is chosen to encode the chosen amino acid sequence. For example, it is common to use a mixture of oligonucleotides as probes for any particular sequence of amino acids, with each oligonucleotide having the same nucleotide base sequence except at specific bases which are varied to take into account the redundancy of the codons that may code for any particular amino acid. It is of course desirable to select an amino acid sequence which is encoded by as few different oligonucleotides as possible. In addition, the various redundant codons may be specifically selected to represent those codons that are most preferred in, for example, hookworm nucleic acid.

In addition, the isolated pure NIF protein can be used to obtain antibodies using known procedures. Such antibodies may include monoclonal or polyclonal antibodies and can be used to screen bacteriophage lambd-agtll expression libraries containing other source (e.g. , hookworm) DNA. In this manner, any particular clone which includes nucleic acid encoding a NIF can be readily identified using standard procedures. Genomic DNA libraries of a hookworm, for example, can be formed using standard procedures to isolate the

genomic DNA of the hookworm, fractionating that DNA using either a random procedure, such as sonication, or a specific procedure such as restriction endonuclease digestion and ligation of those fragments into an appropriate vector, such as a bacteriophage lambda, plasmid or cosmid vector. Such a library can be screened for useful clones by nucleic acid hybridization using the oligonucleotide mixtures described above. More preferably, however, a CDNA library can be constructed by isolation of total hookworm RNA, passage of that RNA over an oligo-dT column to purify the poly(A)-containing RNA (i.e.. messenger RNA), and reverse transcription of such RNA to produce DNA fragments representative of the RNA (i.e.. CDNA) . These CDNA fragments can be inserted using standard procedures into any desired vector, for example, an expression vector such as a commercially available ______ coli expression vector such as bacteriophage lambda-gtll (for expression in E___ coli) , or into a plasmid pcDNA-1 which can be expressed in mammalian COS7 cells.

The biological activity of the protein expressed by individual clones of the plasmid expression library can be readily assayed using the neutrophil inhibitory activity assays described herein or other suitable assays. Alternatively, the antibodies described above can be used to probe for immunoreactive protein expressed from clones in the bacteriophage expression libraries (e.g., lambda-gtll). It is particularly preferred to screen various libraries in sub-pools, for example of 999 clones at a time, to determine which of those sub-pools includes a positive clone. When a positive clone is isolated a grid of the 999 colonies can be formed on a 33 x 33 plate and eaέh of the 33 clones in each row and column in the plate assayed simultaneously (i.e.. in 66 preparations) to identify the desired clone.

Once the desired clone is isolated, its structure is analyzed by standard procedures, for example, by DNA sequencing to determine whether it encodes the whole of the desired protein. If it does not, that clone can be used to screen further CDNA or genomic libraries for full-length clones, or the DNA can be used to hybrid select RNA present in the hookworm, or other source, and more selective CDNA libraries formed from that RNA using procedures described above. In another preferred aspect, the present invention is directed to NIFs comprising an amino acid sequence which is encoded by a nucleic acid sequence which is sufficiently complementary to hybridize to CDNA probes derived from the amino acid sequence of a NIF. Preferred are probes having at least about 12 nucleotides which are complementary to a portion of the sequence of Figure 8.

It should be apparent to those skilled in the art that the oligonucleotide primers can be used in the polymerase chain reaction (PCR) to generate complementary DNA probes. These probes can be used to isolate NIF from other sources or isoforms from a single source. Preferred are animal, fungal, bacterial or viral sources. In the PCR cloning method, single stranded DNA primers of 20-100 nucleotides are derived from the sequence of NIF from Ancylostoma canium. Preferred primers have the following characteristics: limited degeneracy; adherence to codon usage preferences of the particular species from which the library is constructed and primers that target sequences which are conserved among the twelve Ancylostoma NIF isoforms. Each PCR reaction utilizes two primers: a 5-primer that corresponds to the sense strand and a 3^-primer that corresponds to the antisense strand of the NIF coding sequence. Especially preferred are the primers

5-CTCGAATTCT(GATC)GC(ATC)AT(ATC) (CT)T(GATC)GG(ATC)TGGGC- 3' and

5'-CTCGAATTCTT(TC)TCTGG(GA) A(GA)CG(GA)TC(GA) A-3' . The nucleotides within enclosing parentheses are redundant in that any one nucleotide may be used at the position enclosed by such parentheses.

Single stranded CDNA template is generated using poly(A) ⁺ or total RNA prepared from cells of the tissue or organism to be screened. RNA is primed with either random hexanucleotides or oligo d(T) and extended with reverse transcriptase. This reaction product is amplified using an appropriate DNA polymerase (e.g., Taq polymerase) , with a sense and antisense primer, with an appropriate thermocycler. A wide variety of polymerase chain reactj-on conditions may be employed, but initial experiments preferably involve relatively low stringency annealing and elongation steps. Preferred conditions are: cycles 1-3, denaturation at 94°C for 1 minute, annealing at 37°C for l minute and elongation at 72°C for two minutes. The ramp time between annealing and elongation steps is extended to at least 2 minutes for these cycles; cycles 4-40, denaturation at 94°C for 1 minute, annealing at 45°C for 1 minute and elongation at 72°C for two minutes. In subsequent experiments, annealing temperature is increased until a single product results from amplification with each primer pair.

Amplification products from individual amplification reactions are used as hybridization probes to screen genomic DNA or CDNA libraries constructed from the tissue from which PCR was effected. DNA or CDNA from any recombinant plaque or colony that hybridized to these amplification products is selected'' for further analyses. NIF complementary DNAs isolated using the techniques described above are subjected to nucleotide

sequence analysis using the procedure of dideoxy sequencing (Sanger et al, 1977, Proc. Natl. Acad. Sci USA 24:5463-5467) .

NIF CDNA isolates containing open reading frames (i.e.. initiating with a methionine and terminating with a TAA, TGA or TAG stop codon) are inserted into suitable vectors for protein expression in either bacterial, yeast, insect or mammalian cells. Expression systems comprise vectors designed to secrete recombinant protein (i.e., fusion of CDNA isolate open reading frame with a known secretion signal sequence for that cell type) into the culture medium. Vectors lacking a secretion signal sequence are also used for expression. Either conditioned media or cell lysate, depending on the expression system used, is tested for inhibitory activity using one or more of the following criteria for neutrophil activation: release of hydrogen peroxide, release of superoxide anion, release of myeloperoxidase, release of elastase, homotypic neutrophil aggregation, adhesion to plastic surfaces, adhesion to vascular endothelial cells, chemotaxis, transmigration across a monolayer of endothelial cells and phagocytosis.

As discussed above and as described in Example 10, oligonucleotide primers derived from the peptide sequences of NIF (isolated from the hookworm,

Ancylostoma caninum) were used in conjunction with the polymerase chain reaction to amplify NIF CDNA sequences. These NIF sequences were used in turn to probe a hookworm CDNA library. Ten full-length and six partial clone isoforms of NIF were isolated in addition to the protypical NIF-IFL full-length clone. This example illustrates the utility of this technique for isolation of sequences that are structurally related to NIF.

Applicants note that by using techniques such as those described above, as well as similar and equivalent techniques, DNA sequences which encode NIF from other

animal, fungal, bacterial or viral source may be isolated and used to express recombinant NIF.

Should immunoreactive material be expressed from an expression library, the expression vectors described above, or derivatives thereof, can be used for expression of recombinant protein with biological activity. Such recombinant protein is useful in this invention.

Using one example of a NIF of the present invention, peptide fragments were produced and their amino acid sequences determined. This experiment is described in Example 9. The amino acid sequences obtained for the proteolytic fragments are set forth in Figure 7. An example of NIFs being cloned from a canine hookworm CDNA library as described in Examples 10 and 21. The nucleotide sequence for the CDNA of one of the isolated clones (clone 1FL) is depicted in Figure 8. Deduced partial amino acid sequences for other isolated NIF isoform clones are depicted in Figures 9 and 16.

By using the techniques described herein and other techniques in the art, NIFs may be isolated from any source, whether, animal, bacterial, fungal, viral or other source suspected of having a NIF. Such NIFs and nucleic acid sequences encoding them may be isolated by methods such as probing a genomic or CDNA library from the source suspected of having a NIF using oligonucleotide probes sufficiently complementary to a nucleic acid sequence encoding a NIF such as those sequences depicted in Figure 8, and then isolating and expressing those nucleic acid sequences which hybridize to the probes as described herein. Such probes have a sufficient number of nucleotides to desβribe a unique sequence. Typically such probes will have at least about 12 nucleotides. One preferred group of probes include those of the sequences:

5'-CTCGAATTCT(GATC)GC(ATC)AT(ATC)-(CT)T(GATC)GG(ATC)TGGG C-3' and 5'-CTCGAATTCTT(TC)TC-

TGG(GA)AA(GA)CG(GA)TC(GA)AA-3'. The nucleotides within enclosing parentheses are redundant in that any one nucleotide may be used at the position enclosed by such parentheses.

Alternatively, NIF proteins and nucleic acids coding for such proteins may be isolated by probing a sample of nucleic acid from a source suspected of having a NIF with an oligonucleotide probe having at least about 12 nucleotides which is complementary to a nucleic acid sequence known to encode a NIF, such as the sequence depicted in Figure 8 and isolating those nucleic acid sequences, such as a gene, which are sufficiently complementary to the oligonucleotide probe to hybridize thereto. The isolated nucleic acid sequence may then be cloned and expressed using art techniques.

(D) Isolated Nucleic Acid Molecules. In another aspect, the present invention is directed to isolated nucleic acid molecules comprising a nucleic acid sequence encoding the amino acid sequence of NIF. The DNA isolate may also include additional sequences which do not code for portions of the finished protein, such as introns, and/or sequences which code for intervening amino acid residues or peptides in addition to the above peptide sequences. Preferred isolated nucleic acid molecules are those which encode a NIF comprising at least one amino acid sequence selected from the group consisting of

(a) Arg-X _! -X ₂ -Phe-Leu-X ₃ -X ₄ -His-Asn-Gly-Tyr-Arg-Ser- X ₅ -Leu-Ala-Leu-Gly-His-X ₆ -X ₇ -Ile, is Leu or Arg; X ₂ is Gin, Lys or Arg; X ₃ is Ala or Arg; X ₄ is Leu or Met; X ₅ is Lys, Arg, Leu or lie; X ₆ is Val or lie; and X ₇ is Ser, Gly or Asn;

(b) Ala-X ₈ -X ₉ -Ala-Ser-X ₁₀ -Met-Arg-X _n -Leu-X ₁₂ -Tyr- Asp-Cys-Xι ₃ -Ala-Glu-Xι ₄ -Ser-Ala-Tyr-X ₁₅ -Ser-Ala, wherein X ₈ is His or Pro; X, is Thr, Arg or Ser; X ₁₀ is Arg or Lys; X _u is lie or Tyr; X ₁₂ is Asp, Lys or Glu; Xj ₃ is Asp or Glu; X ₁₄ is Gly, Lys or Arg; and X ₁₅ is Glu, Met, Thr or Val;

(c) Ser-X. ₆ -Phe-Ala-Asn-X ₁₇ -Ala-Trp-Asp-X ₁₈ -Arg- Glu-Lys-X ₁₉ -Gly-Cys-Ala-Val-Val-X ₂₀ -Cys, wherein X ₁₆ is Asn or Asp; X_ _η is Val or Leu; X ₁₈ is Ala or Thr; X ₁₉ is Leu, Val or Phe; and X ₂₀ is Thr, Lys or Asn;

(d) His-Val-Val-Cys-His-X ₂₁ -X ₂₂ -Pro-Lys, wherein X ₂₎ is Tyr or lie; X ₂₂ is Gly or no residue;

(e) Ile-Tyr-X ₂₃ -X ₂₄ -Gly-X ₂₅ -Pro-Cys-X ₂₆ -X ₂₇ -Cys-X ₂₈ - X ₂₉ -Tyr, wherein X^ is Thr, Ser, Lys or Glu; X^ is Thr, Val or lie; X^ is Val, Lys or Thr; X ₂₆ is Arg," Ser or

Asp; X ₂₇ is Asn, Gly, Asp or Arg; X ₂₈ is Asn, Ser or Thr; and X ₂₉ is Gly, Glu or Asp; and

(f) Cys-X ₃₀ -X ₃ ι-Asp-X ₃₂ -Gly-Val-Cys-X ₃₃ -Ile, wherein X ₃₀ is His, lie or Asn; X _3l is Ala, Pro or Asp; X ₃₂ is Glu, Val, Asp or lie; and X ₃₃ is lie, Val or Phe. Especially preferred isolated nucleic acid molecules include those wherein its the coding region has the nucleotide sequence and/or codes for a protein having the deduced amino acid sequence set forth in Figure 8.

(E) Expression of NIF.

In another aspect, the present invention is directed to methods for making biologically active NIFs, wherein such NIFs are expressed intracellularly and are optionally secreted; expression vectors encoding NIF; and host cells transformed with these expression vectors which express and, optionally, secrete NIF.

The CDNA encoding NIF may be inserted into a replicable vector for expression, resulting in the synthesis of biologically active recombinant NIF. Many vectors are available for expression of heterologous

proteins and selection of the appropriate vector will depend primarily on the desired properties of the host cell. Each of the available vectors contain various components specific to the host cell to be transformed. The vector components or control elements generally include, but are not limited to, one or more of the following: a signal sequence, an origin of replication, one or more marker genes, a promoter, an enhancer element and a transcription termination sequence. Once the expression vector containing the inhibitor is constructed, a suitable host cell is transfected or transformed with the expression vector, and recombinant NIF is purified either from the host cell itself or the host cell growth medium. In general, the signal sequence may be a-component of the vector, or it may be encoded by the NIF DNA that is inserted into the vector. If the native inhibitory factor is a secreted gene product (i.e., from the hookworm (or other source) cells) , then the native pro-NIF from hookworm DNA may encode a signal sequence at the amino terminus of the polypeptide that is cleaved during post-translational processing of the polypeptide to form the mature NIF.

All vectors contain a nucleic acid sequence that enables the vector to replicate in one or more selected host cells. Generally, in cloning vectors this sequence is one that enables the vector to replicate independently of the host chromosomal DNA, and includes origins of replication or autonomously replicating sequences. Such sequences are well known for a variety of bacterial, yeast, insect and mammalian cells. The origin of replication from the plasmid PBR322 is suitable for most for most gram-negative^ bacteria, the 2m plasmid origin is suitable for yeast, the baculovirus origin is suitable for some insect cells (e.g.. Sf9 cells; ATCC# CRL1711) and various viral origins (e.g. ,

SV40, adenovirus) are useful for cloning vectors in mammalian cells.

Expression vectors preferably contain a selection gene, also termed a selectable marker. This gene encodes a protein necessary for the survival or growth of transformed host cells grown in selective culture medium. Host cells not transformed with the vector containing the selection gene will not survive in the culture medium. Typical selection genes encode proteins that (a) confer resistance to antibiotics or other toxins, e.g., ampicillin, neomycin or methotrexate, (b) complement auxotrophic deficiencies, or (c) supply critical nutrients not available from complex media. Expression vectors contain promoters that are recognized by the host organism. Promoters are untranslated sequences located upstream (5') to the start codon of a structural gene (generally within about 100 to 1000 base pairs) that control the transcription and translation of a particular nucleic acid sequence, such as hookworm NIF, to which they are operably linked.

A large number of promoters recognized by a variety of potential host cells are well known. These promoters are operably linked to DNA encoding the NIF by inserting the latter into the vector in a way such that the 5' terminus of the NIF DNA is in close linear proximity to the promoter.

Transcription of a DNA encoding the NIFs of this invention by higher eukaryotes is often increased by inserting an enhancer sequence into the vector. (For example, see, Kriegler, M. , 1991, Gene Transfer and Expression, pages 4-18, W.H. Freeman, New York). Enhancers are cis-acting elements of DNA, usually about 10-300 base pairs in length, that act on- a promoter to increase its transcription. Enhancers are relatively orientation and position independent. Typically, one will use an enhancer from a eukaryotic cell virus for

expression in mammalian cells. Examples include the SV40 enhancer, the cytomegalovirus early promoter enhancer and the adenovirus enhancers.

Expression vectors used in eukaryotic (i.e., non-bacterial) host cells will also contain sequences necessary for the termination of transcription and for stabilizing the mRNA. Such sequences are commonly available from the 5' end and, occasionally from the 3' untranslated regions of eukaryotic or viral DNAs. Preferred expression vectors of the present invention include but are not limited to PHIL7SP-Nlcl0, PSG5/NIF1FLCR1, Pma5-NIl/3 PAN-NIF-1FL, pYAM7SP-hNIFl/Δ,Gll-5, pYAM7SP-hNIFl/Δ,Gll-5, pYAMSP-AcaNIF4, pYAMSP-AcaNIF6, pYAMSP-AcaNIF9, pYAMSP-AcaNIF24 and pAN-AceNIF3, the construction of which is described in the Examples.

Suitable host cells for the expression vectors described herein include bacterial, yeast, insect or mammalian cells. Preferred bacteria include E^ coli strains, preferred yeast include Saccharomyces cerevisiae and Pichia pastoris. a preferred insect cell line is Sf9 (ATCC# CRL 1711) include preferred mammalian cell lines are COS-7 (ATCC# CRL 1651) , CHO dhfr' (ATCC# CRL 9096) , CHO-K1 (ATCC# CCL 61) and HeLa (ATCC# CCL 2) . These examples of host cells are illustrative rather than limiting. Preferably the host cell should secrete minimal amounts of proteolytic enzymes. Particularly suitable host cells for the expression of glycosylated NIF are derived from multicellular organisms. Such host cells are capable of complex post-translational processing and glycosylation of expressed proteins.

Host cells are transfected and preferably transformed with the above-described expression vectors of this invention and cultured in conventional nutrient media modified as appropriate for inducing promoters and selecting transformants. Transfection refers to the

taking up of an expression vector by a host cell. Numerous methods of transfection are known to the ordinarily skilled artisan, for example, calcium phosphate coprecipitation, spheroplasting transformation and electroporation. Successful transfection is generally recognized when any indication of the operation of this vector occurs within the host cell. Transformation means introducing DNA into an organism so that the DNA is replicable, either as an extrachromosomal element or chromosomal integration.

Depending on the host cell used, transformation is done using standard techniques appropriate to such cells (e.g., calcium chloride or electroporation for bacterial cells; spheroplasting or electroporation for yeast cells; calcium phosphate or electroporation for insect and mammalian cells) .

The recombinant hookworm NIF preferably is recovered from the culture medium as secreted polypeptide, although it may also be recovered from host cell lysates when expressed intracellularly without a signal or secretory sequence. The expressed hookworm NIF may be purified from culture medium or from cell lysates by a variety of separation techniques including, but not limited to, gel filtration, affinity and ion exchange chromatography, hydroxyapatite chromatography,

C4 reverse-phase HPLC and preparative isoelectric chromatography.

(F) Mutant NIFs.

In another aspect, the present invention is directed to mutant NIFs comprising the amino acid sequence shown in Figure 8, wherein one or more of asparagine residues at positions 10, 18,-87, 110, 130, 197 or 223 is replaced by an glutamine residue.

Amino acid sequence variants of NIF may be prepared by introducing nucleotide changes into the DNA which

encodes NIF, isolated as described above. Such variants include substitutions of residues within the amino acid sequence of the NIF. Any combination of substitutions can be made to arrive at the final construct, provided that the final construct possesses certain desired characteristics. The desired characteristics include, but are not limited to, an increased potency over the wild-type NIF and alteration of the amount of glycosylation of NIFs. Once variant NIF DNAs have been constructed, variant recombinant forms of NIF may be synthesized utilizing expression systems as described above.

One possible method for preparing variants of NIF is mutagenesis with base-specific chemical mutagens as described in detail by Pine and Huang (1987, Methods Enzymol. 154. 415-430) . Another approach is site-directed mutagenesis, for example, as described in Stanssens et al. (1989), Nucl. Acids Res., 17: 4441-4454. For example. Example 25 describes the substitution of certain asparagine residues in the amino acid sequence of a NIF referred to as IFL (see Figure 8) by site-specific mutagenesis.

The asparagine residues at positions 10, 18, 87, 110, 130, 197 and 223 of the amino acid sequence of NIF isoform, IFL, are believed to be sites associated with the potential N-linked glycosylation of this isoform. Using the stepwise site-directed procedure of Stanssens et al. and the five oligonucleotides described in Example 25, asparagine residues at positions 10, 18, 87, 110 and 130 of IFL were replaced by glutamine residues to yield a mutant NIF. Subsequently, the cDNA encoding this mutant was further mutagenized by the same procedure using other oligonucleotides described to produce a mutant NIF, wherein the asparagine residues at positions 197 and 223 of the amino acid sequence of IFL were also replaced with glutamine residues. In either

case, the expressed mutant NIFs were found to have neutrophil inhibitory activity.

2. Peptide Fragments.

In another aspect, the present invention is directed to peptide fragments having neutrophil inhibitory activity which are prepared by proteolytic or chemical methods starting with the chromatographically pure NIF of the present invention.

Active peptide fragments, with or without sugar moieties, may be generated by using enzymatic or chemical techniques. Proteolytic cleavage can be accomplished by digestion of the inhibitor with one or more of the following enzymes: chymotrypsin, trypsin, leucine aminopeptidase, endoproteinase Glu-C, - endoproteinase Lys-C, endoproteinase Arg-C, or endoproteinase Asp-N (Carrey, E.A. , 1989 Protein Structure. A Practical Approach, pp. 117-143, T.E. Creighton, ed. IRL Press, New York) . Chemical digestion of the inhibitor may be accomplished by cyanogen bromide, hydroxylamine, or 2-nitro-5-thiocyanobenzoate cleavage (Carrey, E.A., 1989, ibid.) . Sugar moieties can be removed from either the peptide fragments or intact neutrophil inhibitory protein enzymatically with one or more of the following enzymes: glycopeptidase F, endoglycosidase H, endoglycosidase F, or endoglycosidase

D as described by Keesey (Keesey, J. , 1987 Biochemica Information, pp. 147-165, J. Keesey, ed. , Boehringer Mannheim Biochemicals, Indianapolis) . Alternatively, glycosylation of the intact inhibitor may be suppressed by expression of the protein in bacterial cells or by the inclusion of inhibitors of glycosylation in the eukaryotic cell culture growth medium. Inhibitors of glycosylation and their uses are described in the art (e.g., Keesey, J. 1987 Biochemica Information, pp. 135-141, J. Keesey, ed. , Boehringer Mannheim

Biochemicals, Indianapolis) . Separation of active fragments from inactive fragments may be accomplished by conventional, low, medium, or high pressure chromatographic techniques known in the art.

3. Antibodies.

In another aspect, the present invention is directed to polyclonal and monoclonal antibodies which have the ability to bind to NIFs.

To prepare antibodies to NIF, any one of a number of conventional techniques which are known in the art can be employed. In one such technique, polyclonal antibodies are synthesized by injecting an animal (for example a rabbit) with one or more NIF of the present invention. After injection, the animal produces antibodies to these NIFs. When the antibody concentration (or titer) reaches a sufficient level, antibody-containing blood is then drawn from the animal, antiserum is prepared from the blood, and the compound-specific antibody is isolated from other antibodies in the serum by any one of a number of separation techniques (for example, affinity chromatography) .

Monoclonal antibodies may be prepared using the technigues of Kohler and Milstein, Nature 256. 495-497 (1975) as well as other conventional techniques known to those skilled in the art. (See, e.g. , Harlow and Lane, Antibodies. A Laboratory Manual (Cold Spring Harbor Laboratory, 1988) the disclosures of which is incorporated herein by reference) . Preferred monoclonal antibodies include those directed .to the NIF isoform,

IFL, whose amino acid sequence is depicted in Figure 8 and which are immunoglobulins of the IgG*class. Especially preferred monoclonal antibodies are those which bind to the same epitope on this NIF as is bound by the monoclonal antibody, 3D2. The monoclonal

antibody referred to as 3D2 is a preferred embodiment. The preparation of 3D2 is described in Example 26.

In another aspect, the present invention is directed to hybridomas which produce such monoclonal antibodies. These hybridomas are produced by conventional techniques such as those described by Harlow and Lane, Id. , the disclosures of which is incorporated herein by reference. The preparation of a preferred hybrido a is described in Example 26. In another aspect, the present invention is directed to methods of affinity purification of NIF from various sources and impure compositions of NIF derived from such sources using an antibody to NIF. Preferred method of isolating NIF would comprise the step of contacting a sample thought to contain a NIF _^ with a monoclonal antibody which is capable of binding to said NIF. Preferred as the monoclonal antibody is either the monoclonal antibody, 3D2, or a monoclonal antibody binding to the same epitope on said NIF as is bound by the monoclonal antibody, 3D2. According to preferred methods of affinity purification the monoclonal antibody is covalently attached to a chromatographic resin. Especially preferred chromatographic resins include E phaze ⁶¹¹ Biosupport Medium (3M Corp.). Examples 27 and 28 describe preparation of a chromatographic resin coupled with the monoclonal antibody, 3D2, and its use to purify NIF from compositions comprising NIF.

In another aspect, the present invention is directed to immunoassays using the antibodies against NIF. Depending on the particular use for the im unoassay, an immunoassay format is selected. Some suitable immunoassays are described by Harlow and Lane, Id. (See especially pages 553 to 612)^' the disclosures of which are incorporated herein by reference. Immunoassays utilizing the solid phase method or liquid phase method are well known to one skilled in the

art of immunoassays. For example, monoclonal antibodies may be used to assay for drugs, hormones and proteins with such assays being in a solid phase or liquid phase format. The preferred methods of the present invention are solid phase assays.

The preferred methods of detecting NIF in a sample comprise contacting a sample thought to contain a NIF with a monoclonal antibody which is capable of binding to such NIF. According to a preferred embodiment, the monoclonal antibody is immobilized onto a plastic surface such polystyrene, polypropylene, polyethylene, nylon and the like. The plastic surface may be configured in the shape of test tube, microspheres, macroscopic beads, microtiter plates and the like. In this embodiment, monoclonal antibodies may be attached to the plastic surface by either covalent coupling or by passive absorption, preferably by passive absorption. Preferred as monoclonal antibodes are those which bind to the same epitope on a NIF as is bound by the monoclonal antibody, 3D2, or the monoclonal antibody,

3D2.

The monoclonal antibody may be contacted simultaneously with sample and a NIF which has been covalently linked to a detectable label. Alternatively, the monoclonal antibody may first be contacted with the sample, then with a NIF which has been covalently linked to a detectable label.

Preferred detectable labels are enzymes, fluorescent compounds or radioisotopes. Especially, preferred detectable labels are enzymes such as alkaline phosphatase, β-galactosidase or horseradish peroxidase, or radioisotopes such as iodine-125. The manner of covalently linking such enzymes and radioisotopes to monoclonal antibodies is well known to one skilled in the art of diagnostic assays.

4. Methods of Detecting NIF Mimics and NIF Antagonists.

In another aspect, the present invention is directed to a method of detecting in a sample the presence of a NIF mimic which competes with NIF for binding to CDllb/CD18 receptor or the I-domain portion of the CDllb/CD18 receptor. In another aspect, the present invention is directed to a NIF antagonist which prevents NIF from binding to the CDllb/CD18 receptor or the I-domain portion of the CDllb/CD18 receptor. The methods comprise contacting said sample with CDllb/CD18 receptor or the I-domain portion of the CDllb/CD18 receptor. NIF mimics and NIF antagonists include but are not limited to small molecules, peptides, peptide analogs or proteins.

In preferred embodiments, NIF mimics or antagonists to be tested are preincubated in solution with neutrophils, or immobilized CDllb/CD18 receptor or a recombinant peptide comprising the I-domain portion of the CDllb/CD18 receptor, and the preincubated solution is then brought into contact with labeled NIF. The effect of test compound on the binding of NIF to neutrophils, immobilized CDllb/CD18 receptor or recombinant peptide comprising the I-domain portion of the CDllb/CD18 receptor is then determined.

In one preferred embodiment, the assay method uses neutrophils which are free in solution. Here, NIF which has been linked to a detectable label, neutrophils and a sample thought to contain a NIF mimic or NIF antagonist are co-incubated in solution for a sufficient time to allow binding to occur. Unbound labeled NIF is removed from bound NIF by methods such as centrifugation, filtration or other suitable methods and bound NIF is determined by means of the detectable label. In another preferred embodiment, neutrophils are immobilized on a plastic surface by passive absorption

or chemical fixation such as by glutaraldehyde or similar chemicals. Labeled NIF is co-incubated with the immobilized neutrophils and a sample thought to contain a NIF mimic or NIF antagonist. Unbound labeled NIF is removed by washing and bound labeled NIF is then determined by means of the detectable label.

In an especially preferred embodiment, CDllb/CD18 receptors from a detergent extract of human leukocytes are captured by anti-CDllb/CD18 monoclonal antibody which is immobilized to a plastic surface. A NIF which is linked to a detectable label or a NIF which can be subsequently linked to a detectable label, and sample thought to contain a NIF mimic or NIF antagonist are co-incubated with the immobilized CDllb/CD18 receptor. After the binding has occurred, unbound NIF JLs removed by washing. Detectable label is then added which links to NIF rendering it detectable (if the NIF was not originally linked to such label) . Bound NIF is then determined by means of the detectable label. Anti-CDllb/CD18 antibody may be coupled to a plastic surface by covalent coupling or passive absorption, though passive absorption is preferred. Preferred plastic surfaces are polystyrene, polypropylene, polyethylene, nylon and the like, though polystyrene is preferred. The plastic surface may be configured in the shape of test tube, microspheres, macroscopic beads, microtiter plates and the like. A preferred anti-CDllb/CD18 antibody is the monoclonal antibody referred as LM2. NIF is linked to a detectable label or is capable of being linked to such label during a step of an assay method of the present invention. NIF is linked to a detectable label by covalent coupling^ύsing homobifunctional crosslinking reagents such as glutaraldehyde, disuccinimidyl suberate, dimethyl suberimidate and the like. NIF is made capable of being

linked to a detectable label during a step of an assay method of the present invention by first linking biotin to NIF and avidin to the detectable label. Preferred detectable labels include enzymes, fluorophores or radioisotopes. Especially preferred detectable labels include alkaline phosphatase, β-galactosidase, horseradish peroxidase, or iodine-125.

In another especially preferred embodient, a recombinant peptide comprising the I-domain of the CDllb/CD18 receptor which is complexed to a monoclonal antibody that recognizes a portin of this peptide NIF which is linked to a detectable label or a NIF which can be subsequently linked to a detectable label, and sample thought to contain a NIF mimic or NIF antagonist are co- incubated. After the binding has occurred, a .protein A- Sepharose is added to separate bound from unbound NIF. Setectable label is then added which links to NIF rendering it detectable (if the NIF was not originally linked to such label) . Bound NIF is then detrmined by means of the detectable label.

A NIF mimic or NIF antagonist may be assayed for neutrophil inhibitory activity. Neutrophil inhibiting activity may be demonstrated by an assay such as those assays which determine adhesion of neutrophils to vascular endothelial cells, release of hydrogen peroxide from neutrophils, homotypic neutrophil aggregation and adhesion of neutrophils to plastic surfaces. NIF mimics are characterized as having an IC ₅₀ for inhibiting neutrophil activity of about 500 nM or less, though an IC _J0 is about 100 nM or less is preferred. NIF antagonists are characterized having such an IC ₅₀ of about 1,000 nM to about ImM, although an IC _J0 of about 5,000 nM to about 10 mM is preferred. An IC _j0 is that concentration of a NIF mimic or NIF antagonist giving 50% inhibition of the measured activity (see Example l) .

Optionally a NIF mimic or NIF antagonist may be assayed for eosinophil inhibitory activity. Eosinophil inhibiting activity is demonstrated by an assay which determines adhesion of eosinophils to vascular endothelial cells. If assayed, NIF mimics are characterized as having an IC ₅₀ for inhibiting eosinophil activity of about 500 nM or less, though a IC ₅₀ is about 100 nM or less is preferred. If assayed, NIF antagonists are characterized having such an IC ₅₀ of about 1,000 nM to about 1 mM, though an IC _J0 of about 5,000 nM to about 10 mM is preferred. An IC ₅₀ is that concentration of a NIF mimic or NIF antagonist giving 50% inhibition of the measured activity.

In another aspect, the present invention is directed to NIF mimics and NIF antagonists discovered by the above-disclosed methods of detection of this section.

5- Pharmaceutical Formulations and Methods of Treatment. In another aspect, the present invention is directed to pharmaceutical compositions comprising NIF.

These pharmaceutical compositions may be formulated and used as tablets, capsules or elixirs for oral administration; suppositories for rectal administration; sterile solutions, suspensions for injectable administration; and the like. The dose and method of administration can be tailored to achieve optimal efficacy but will depend on such factors as weight, diet, concurrent medication and other factors which those skilled in the medical arts will recognize.

Generally, an amount between 0.01 mg/kg to 100 mg/kg body weight/day is administered dependent upon the potency of the composition used.

Preferred embodiments encompass pharmaceutical compositions prepared for storage and subsequent

administration which comprise a therapeutically effective amount of NIF or an enriched composition of NIF, as described herein in a pharmaceutically acceptable carrier or diluent. Acceptable carriers or diluents for therapeutic use are well known in the pharmaceutical art, and are described, for example, in Remington's Pharmaceutical Sciences. Mack Publishing Co. (A.R. Gennaro edit. 1985). Preservatives, stabilizers, dyes and even flavoring agents may be provided in the pharmaceutical composition. For example, sodium benzoate, sorbic acid and esters of p-hydroxybenzoic acid may be added as preservatives. In addition, antioxidants and suspending agents may be used. In another aspect, the present invention is directed to methods of preventing in a mammal an inflammatory condition characterized by abnormal neutrophil activation or abnormal eosinophil activation comprising administering to said mammal a therapeutically effective amount of a NIF or their pharmaceutical compositions. In practicing the preferred methods, NIFs or their pharmaceutical compositions can be used alone or in combination with one another, or in combination with other therapeutic or diagnostic agents. These compositions can be utilized in vivo, ordinarily in a mammal, preferably in a human, or in vitro.

In employing NIFs or their pharmaceutical compositions in vivo, the compositions can be administered to the mammal in a variety of ways, including parenterally, intravenously, subcutaneously, intramuscularly, colonically, rectally, nasally or intraperitoneally, employing a variety of dosage forms. As will be readily apparent to one skilled in the art, the useful in vivo dosage to be administered and the particular mode of administration will vary depending upon the mammalian species treated, the particular

composition employed, and the specific use for which these compositions are employed. The determination of effective dosage levels, that is the dosage levels necessary to achieve the desired result, will be within the ambit of one skilled in the art. Typically, applications of compositions are commenced at lower dosage levels, with dosage level being increased until the desired effect is achieved.

The dosage for a NIF or its pharmaceutical compositions can range broadly depending upon the desired effects and the therapeutic indication. Typically, suitable dosages will be between about 0.01 mg and 100 mg/kg, preferably between about 0.01 and 10 mg/kg, body weight. Administration is preferably parenteral, such as intravenous on a daily or as-needed basis.

Injectables can be prepared in conventional forms, either as liquid solutions or suspensions, solid forms suitable for solution or suspension in liquid prior to injection, or as emulsions. Suitable excipients are, for example, water, saline, dextrose, mannitol, lactose, lecithin, albumin, sodium glutamate, cysteine hydrochloride or the like. In addition, if desired, the injectable pharmaceutical compositions may contain minor amounts of nontoxic auxiliary substances, such as wetting agents, pH buffering agents, and the like. If desired, absorption enhancing preparations (e.g., liposomes) may be utilized.

6. Utility and Applications As applicant has noted above, the present invention is directed to inhibitors of the leukocytes, and in a preferred aspect, to those leukocytes which have or express CDllb/CD18 receptors. Leukocytes having or expressing the CDllb/CD18 integrin receptor are well known and include the monocytes, macrophages,

granulocytes, large granular lymphocytes (NK cells) , and immature and CD5 ⁺ B cells (Kishimoto, T.K. , Larson, R.S., Corbi, A.L., Dustin, M.L., Staunton, D.E., and Spriger, T.A. (1989) Adv. in Immunol. 46,149-182). CDllb/CD18 has been implicated in a variety of leukocyte functions including adhesion of neutrophils to endothelial cells (Prieto, J. , Beatty, P.G., Clark, E.A., and Patarroyo, M. (1988) Immunology 63, 631-637; Wallis, W.J., Hickstein, D.D., Schwartz, B.R., June, C.H., Ochs, H.D., Beatty, P.G., Klebanoff, S.J., and Harlan, J.M. (1986) Blood 67, 1007-1013; Smith, C.W. , Marlin, S.D., Rothlein, R. , Toman, C, and Anderson, D.C. (1989) J. Clin. Invest. 83, 2008-2017) and release of hydrogen peroxide from neutrophils (Shappell, S.B., Toman, C, Anderson, D.C, Taylor, A.A. , Entman, M.L. and Smith, C.W. (1990) J. Immunol. 144, 2702-2711; Von Asmuth, E.J.U. , Van Der Linden, C.J. , Leeuwenberg, J.F.M., and Buurman, W.A. (1991) J. Immunol. 147,3869- 3875) . This integrin may play a roll in neutrophil and monocye phagocytosis of opsonized (ie C3bi-coated) targets (Beller, D.I., Springer, T.A. , and Schreiber, R.D. (1982) J.Exp. Med. 156,1000-1009). It has also been reported that CDllb/CD18 contributes to elevated natural killer activity against C3bi-coated target cells (Ramos, O.F., Kai, C, Yefenof, E., and Klein, E. (1988) J. Immunol. 140,1239-1243).

Also noted previously, the NIFs of the present invention have potent neutrophil inhibitory activity and, thus, may be used as an inhibitors of neutrophil activity, including neutrophil activation in vitro, as well as for preventing or treating in a mammal inflammatory conditions characterized by abnormal neutrophil activation. * ^•

Thus, NIF will be useful in the treatment of inflammation in which the abnormal activation of neutrophils plays a significant role. While applicants

do not wish to be bound to any theory or mode of activity, it is believed that this compound will interfere with the inflammatory response which is set into action by neutrophil-endothelial cell interactions. Thus, where adhesion of neutrophils to the endothelium is prevented, the neutrophils will be unable to transmigrate to tissue to elicit a proinflammatory response with consequent tissue damage. Inhibition of neutrophil-neutrophil adhesion and/or aggregation by these NIFs should also prevent microvascular occlusion. Thus, these NIFs will be useful in treating a variety of clinical disorders, including shock, stroke, acute and chronic allograft rejection, vasculitis, autoimmune diabetes, rheumatoid arthritis, inflammatory skin diseases, inflammatory bowel disease, adult respiratory distress syndrome (ARDS) , ischemia-reperfusion injury following myocardial infarction, in which neutrophil infiltration and activation has been implicated and acute inflammation caused by bacterial infection, such as sepsis or bacterial meningitis.

The ability of NIF to inhibit neutrophil activity makes it useful in inhibiting the physiological processes of inflammation, ischemia, and other neutrophil mediated tissue damage. The specific activities of NIFs in carrying out these related functions makes it particularly useful as therapeutic and/or diagnostic agents.

Antibodies, both monoclonal and polyclonal, directed to NIF are useful for diagnostic purposes and for the identification of concentration levels of the subject peptides in various biological fluids. Immunoassay utilizing these antibodies may be used as a diagnostic test, such as to detect infection of a mammalian host by a parasitic worm or to detect NIF from a parasitic worm in a tissue of the mammalian host.

Also such immunoassays may be used in the detection and

isolation of NIF from tissue homogenates, cloned cells and the like.

In another aspect of the present invention, NIFs can be used in a test method to screen other compounds to detect NIF mimics or to detect NIF antagonists for their ability to affect NIF binding to the CDllb/CD18 receptor.

In yet another aspect of the present invention, NIF with suitable adjuvants can be used as a vaccine against parasitic worm infections in mammals. Immunization with NIF vaccine may be used in both the prophylaxis and therapy of parasitic infections. NIF fragments and synthetic polypeptides having the amino acid sequence of NIF may also be used as vaccines. Disease conditions caused by parasitic worms may be treated by - administering to an animal infested with these parasites substances which antagonize NIF (such as NIF antagonists) . Compounds may be screened for their anti-NIF effect according to the screening method described herein above. Examples of such antihelminic agents include antibodies to NIF, both naturally occurring antibodies isolated from serum and polyclonal and monoclonal antibodies described above. Chemically synthesized compounds which act as inhibitors of NIF also are suitable antihelminic agents.

To assist in understanding the present invention, the following examples are included which describe the results of a series of experiments. The following examples relating to this invention should not, of course, be construed as specifically limiting the invention and such variations of the invention, now known or later developed, which would be within the purview of one skilled in the art are considered to fall within the scope of the invention as described herein and hereinafter claimed.

Examples

Example 1

Assays of Neutrophil Inhibitory Activity

The Neutrophil Inhibitory Factor of the present invention demonstrated activity in inhibiting neutrophil function as measured by neutrophil-HUVEC and neutrophil-plastic adhesion assays, homotypic neutrophil aggregation assay and hydrogen peroxide release assay. This inhibitory factor was isolated from hookworm tissue lysates as an enriched composition by a variety of methods including gel filtration chromatography, chromatography on hydroxyapatite and concanavalin A sepharose, C4 reverse-phase HPLC, Mono-Q ion exchange chromatography and preparative isoelectric focusing. The isolated factor appears to inhibit neutrophil adhesion to endothelial cell monolayers by inhibiting neutrophil activation.

(A) Cells and Reagents

Primary human umbilical vein endothelial cells (HUVEC) , obtained from Clonetics (San Diego, CA) , were maintained in EGM-UV medium (Clonetics) with 15% fetal bovine serum (FBS) , in a 5% C0 ₂ atmosphere. HUVEC were passaged twice and used to seed fibronectin-coated 96 well microtiter plates (Collaborative Research, Bedford, MA) for adhesion assays.

The protease inhibitors E64, pepstatin A, chymostatin and APMSF were obtained from Calbiochem (La Jolla, CA) .

Neutrophils were isolated using Mono-Poly resolving medium (ICN Biomedicals, Costa Mesa, CA) from either heparinized or citrated human blood following the instructions of the manufacturer. Neutrophils were resuspended in HSA buffer (RPMI1640 with 10 mM HEPES pH 7.4, 1.2 mM CaCl, 1.0 mM MgCl, 1% human serum albumin)

at a concentration of 6.6xl0 ⁶ cells/ml and used within one hour after isolation.

Neutrophils were fluorescently labelled by the following procedure. The cells were washed once in Hank's balanced salt solution (HBSS) and resuspended at lxlO ⁷ cells/ml in HBSS containing 20 mg/ml calcein (Molecular Probes; Eugene, OR) . The calcein was initially solubilized in 50 ml dry dimethylsulfoxide prior to its addition to the HBSS. Cells were incubated at 37°C with occasional mixing by inversion. After 45 minutes incubation the cells were chilled on ice for 5 minutes and then washed twice with ice-cold HSA buffer. Labelled neutrophils were resuspended in HSA buffer at 1.3xl0 ⁷ cells/ml for use in adhesion assays.

(1) Protein concentration.

The molar protein concentration of purified NIF isoforms and mutants thereof was determined spectrophotometrically at 278 nm thereby using calculated extinction coefficients. The calculation is based on absorbance values of 5600 cm" ¹ .mol" ¹ for tryptophan and 1420 cm' ¹ .mol" ¹ for tyrosine residues.

Two hundred fifty microliters glutaraldehyde (25% solution) was added to 1 ml of an 8 mg/ml solution of horseradish peroxidase (Sigma P8375; in phosphate buffered saline (PBS)). After 2 hours at room temperature the mixture was transferred to a dialysis bag (MW cut-off 12kD) and dialysed against 1 liter PBS for 5 hours. Buffer was refreshed three times. After

transfer of the solution to a test tube an equal volume of 3D2 purified MAb (see Example 26; dissolved in PBS at a concentration of 2mg/ml) was added and further incubated overnight at room temperature. The reaction was stopped by adding glycine to a final concentration of 50 mM. Bovine serum albumin (BSA; Calbiochem) was added to a final concentration of 0.25% (w/v) and the solution was aliquoted and stored at -20'C.

(3) Preparation of biotinylated Pichia NIF-IFL. Purified recombinant NIF (NIF-IFL) from Pichia was dialyzed against 100 mM acetate buffer pH 5.5 at a concentration of lmg/ml. An equal volume of cold 20 mM sodium-metaperiodate in 100 mM acetate buffer pH 5.5 was added. The oxidation reaction was allowed to. proceed for 20 minutes in the dark on ice. The reaction was stopped by adding glycerol to reach a final concentration of 15 mM. The sample was desalted by ultrafiltration on Centricon 30. Biotin-LC hydrazide (Pierce) dissolved in di ethylsulfoxide was added to reach a final concentration of 5 mM and was then further incubated for

2 hrs at room temperature. The biotinylated sample was subsequently dialyzed against PBS using a Spectrapor membrane with a cut-off of 12,000.

(B) Neutrophil-HUVEC Adhesion Assays Calcein-labelled neutrophils (175 ml at 1.32xl0 ⁷ cells/ml) were preincubated for 10 minutes at room temperature with 175 ml of test fraction (diluted in HSA buffer) in the presence of 160 nM phorbol 12-myristate 13-acetate (PMA; Sigma, St. Louis, MO) . PMA is solubilized in dimethylsulfoxide at a stock concentration of 1.6 mM. A 96 well p]*ate was used for this assay. One hundred microliters of this suspension was then aliquoted into each of three replicate wells that contained HUVEC monolayers. Neutrophils were

incubated with the HUVEC monolayer for 30 minutes at 37°C To remove non-adherent cells, wells were first filled with 250 ml HSA buffer, sealed with parafilm and then centrifuged inverted for 3 minutes at 75 X g. Inverted plates were then placed on a rocking platform shaker for 5 minutes, after which contents were decanted off and wells were washed twice with 100 ml HSA buffer. Adherent neutrophils were lysed in 100 ml 0.1% (v/v) Triton X-100 (in 50 mM Tris HCl pH 7.4), and agitated for 10 minutes on a plate shaker. Twenty five microliters of the neutrophil/endothelial cell lysate was transferred to a 96 well microtiter plate that contained 100 ml of 50 mM Tris pH 7.4, and the wells were read at 530 nm (485 nm excitation) on a Cytofluor fluorometric plate reader (Millipore; Bedford, MA) . The hydroxyapatite pool preparation of hookworm Neutrophil Inhibitory Factor (see Example 1(D)) inhibited neutrophil adhesion to HUVEC monolayers with an IC ₅₀ of about 10 nM.

(C) Neutrophil-Plastic Adhesion Assay (1) Protocol #1.

This assay was used in Examples 1 through 19, where it was applicable.

Neutrophils (20 ml at 6.6X10 ⁶ cells/ml) were incubated with 5 ml PMA (0.8 mM) for 5 minutes at room temperature in a 0.5 ml polypropylene test tube. Twenty microliters of test fraction, diluted in HSA buffer, was added and the suspension was mixed gently. Aliquots of 10 ml of this suspension were added in triplicate to microtiter wells of 60-well HCA (Terasaki) plates (Nunc, Naperville, IL) . Neutrophils were incubated 5 minutes at 37°C and non-adherent cells were removed by submerging the plate 6 times in HBSS.

Adherent neutrophils were quantitated by counting under an inverted light microscope. Binding was

quantitated visually. PMA-activated neutrophils spread and adhere tightly to polystyrene plastic. Non-activated neutrophils (i.e., in the absence of PMA) remain round and translucent and do not adhere tightly to plastic. Adherent neutrophils were larger, rhomboid in shape and more opaque, with a granular appearance. In the absence of Neutrophil Inhibitory Factor, greater than 80% of PMA-activated neutrophils rapidly and irreversibly bound plastic, underwent shape change and were not removed by the gentle wash procedure. Moreover, fractions containing the Ancylostoma Neutrophil Inhibitory Factor exhibited a profound inhibitory effect on plastic binding by activated neutrophils. The hydroxyapatite pool preparation of hookworm Neutrophil Inhibitory Factor (see Example 1(D)) inhibited neutrophil adhesion to plastic in this assay with an IC ₅₀ of about 10 nM.

(2) Protocol #2. This assay was used in Examples 20 through 29, where it was applicable.

Neutrophils (60 μl at 8xl0 ⁶ cells/ml- in HSA buffer) were incubated with 15 μl PMA (2.4 mM) and 5 μl CaCl2 0.05 M. After gently mixing 80 μl of the stimulated cell suspension were added to 96-well high binding polystyrene microtiter plates (Costar) containing 20 μl of the test sample diluted in HSA buffer. After 45 minutes incubation at 37°C, non-adherent cells were removed by submerging the plates 4 times in PBS. Adherent cells were loaded with dye by adding lOOμl of

0.5% (w/v) Crystal violet indicator (CAS 548-62-9) solution. After 10 min at room temperature plates were rinsed by submerging 4 times in PBS. Lysis of stained adherent cells was done by adding lOOμl of 1% (v/v) Triton X-100 solution. Absorbance was determined at

605nm with a Thermomax plate reader to quantitate adherent neutrophils.

(D) Homotypic Neutrophil Aggregation

Neutrophil aggregation was performed at 37°C in a Scienco dual channel aggregometer (Morrison, CO) .

Neutrophils (190 ml at 6.6X10 ⁶ cells) were preincubated with 200 ml test fraction (diluted in HSA Buffer) in a glass cuvette (Scienco) for 2 minutes at room temperature. Ten microliters of PMA were added to initiate aggregation (80 nM final) . The inhibition of neutrophil aggregation was measured at the maximum aggregation response 5 minutes after the addition of PMA.

The hydroxyapatite pool preparation of. Neutrophil Inhibitory Factor (see Example 1(D)) inhibited neutrophil adhesion with an IC ₅₀ of about 10 nM.

(E) Hydrogen Peroxide Release Assay

Neutrophils (6.6X10 ⁶ cells/ml) were incubated with test fractions in Release Assay Buffer (HBSS with 25 mM glucose, 10% FBS, 200 mg/ml phenol red, 32 mg/ml horseradish peroxidase) for 5 minutes at 37°C Incubation vessels consisted of 1.5 ml plastic test tubes that were precoated with HBSS containing 50% FBS at 37°C for 60 minutes; coated tubes were washed twice with 0.15 M NaCl before use. FM1P (Sigma; St. Louis,

MO) at a final concentration of 250 mM was added and the neutrophil/test compound suspension was incubated at 37°C for 60 minutes. Cells were pelleted by centrifugation at 8000 X g for 3 minutes and 200 ml of supernatant was transferred to a 96 well microtiter plate. Ten microliters of 1 N NaOH was added to each well and absorbance was read at 610 nm with a Molecular Devices ThermoMax plate reader. Hydrogen peroxide

concentrations were determined by using a standard curve. Data points were done in duplicate.

The hydroxyapatite pool preparation of hookworm Neutrophil Inhibitory Factor inhibited hydrogen peroxide release from neutrophils with an IC ₅₀ of about 10 nM.

(F) CDllb/CD18 Binding Assays

Microtiter polystyrene plates (Costar - high binding; 96 well) were coated with the CDllb/CD18 binding, non-neutralizing mouse monoclonal antibody LM2 (50 μl of the purified LM2 MAb at a concentration of 10 μg/ml in 0.1 M NaHC0 ₃ ; pH 9.5; LM2 ATCC hybridoma #HB204) by overnight incubation at 4°C After removal of the antibody, the wells were blocked with 200 μl PBS containing 1% (w/v) Skim-milk (Difco Laboratories) at room temperature. After 2 hours the blocking solution was removed and wells were washed 3 times with 200 μl of PBS.

The immobilized LM2 monoclonal antibody was used to immuno-capture the detergent solubilized CDllb/CD18 receptor as follows. Neutrophils were isolated using Polymorphprep™ (Nycomed) from either citrated whole blood or from buffy-coat following the instructions of the manufacturer. The neutrophil pellet from a 50 ml buffy coat was resuspended in 40 ml RPMI and phorbol myristate acetate (PMA) to a final concentration of 0.8 μM was added. The suspension was gently rotated at room temperature for 20 minutes. The cells were spun down and resuspended in 5 ml 0.02 M Tris-HCl, 0.15 M NaCl, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ . Cells were lysed by adding 5 ml buffer containing 2% Triton χ-100 (Bio-Rad

Laboratories), 0.02 M Tris pH 7.5, 0.15 M NaCl, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ and 0.02% thimerosa (Sigma T-5125) . PMSF (Sigma) and iodoacetamide (Merck-Schuchardt) were added to final concentrations of 1 mM. After mixing, the suspension was stored on ice for 1 hour and vortexed

periodically. After centrifugation at 30,000 g, the supernatant was diluted twofold with 0.02 M Tris-HCl, 0.15 M NaCl, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ and 2 ml of IgG-sepharose (Sigma) was added. After incubation for two hours in the cold, the resin was spinned down. The supernatant was removed and diluted 5-fold with 0.02 M Tris-HCl, 0.15 M NaCl, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ . Fifty microliters of this neutrophil lysate was then added to LM2-coated wells and incubated for 2 hours at room temperature to capture the CDllb/CD18 integrin.

After washing with PBS, the plates were stored at -20°C

In one type of assay, binding of NIF (NIF-IFL and isoforms or engineered variants) to LM2/CDllb/CD18 coated plates was detected with the 3D2-HRP cpnjugate. Samples containing NIF were diluted in PBS containing

0.1% (w/v) Skim-milk, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ . After

2 hours incubation at room temperature, the wells were washed three times with 200 μl PBS containing 0.001 M MgCl ₂ , 0.001 M CaCl ₂ , 0.02% Tween-20 and 0.02% thimerosal. Then, 100 μl of an appropriate dilution

(typically 2000-fold; in PBS containing 0.1% (w/v) skim-milk, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ ) of the 3D2-HRP conjugate was added. After 1 hour, the wells were washed three times with 200 μl PBS containing 0.001 M MgCl ₂ , 0.001 M CaCl ₂ , 0.02% thimerosal and 0.02% Tween 20. Substrate for HRP (100 μl; prepared by dissolving 2.3 mg ortho-phenylenediamine in 11.6 ml 50 mM citrate, 100 mM disodium phosphate buffer pH 5 and adding 2 μl of

35% H ₂ 0 ₂ ) was then added to the wells. The reaction was stopped (typically after 20 minutes) by addition of 10 μl of a 4 M sulfuric acid solution; the absorbance was read at 605 nm with a Thermomax plate reader (Molecular Devices) .

Alternatively, binding of NIF (NIF-IFL, other NIF proteins, and NIF mutants) to the LM2/CDllb/CD18 complex

was measured by competition with biotinylated recombinant NIF-IFL produced in Pichia (see Example 12) . Samples containing a constant amount of biotinylated recombinant NIF-IFL and varying amounts of unlabeled recombinant NIF-IFL were prepared in PBS containing

0.001 M MgCl ₂ , 0.001 M CaCl ₂ and 0.1% (w/v) casein (Difco Laboratories) . A 100 μl aliquot of each sample was added to individual LM2/CDllb/CD18 coated wells and incubated for 2 hours at room temperature. Unbound material was removed by washing three times with 200 μl PBS containing 0.001 M CaCl ₂ , 0.001 M MgCl ₂ , 0.1% Tween 20 and 0.02% thimerosal. Retained biotinylated recombinant NIF-IFL was detected by incubating first with 100 μl of ExtrAvidin-phosphatase (Sigma) conjugate diluted in PBS containing 0.1% casein. After 1 hour incubation at room temperature, unbound conjugate was removed by washing three times with PBS containing 0.001 M CaCl ₂ , 0.001 M MgCl ₂ , 0.02% Tween 20 and 0.02% thimerosal. Substrate for alkaline phosphatase (100 μl of a solution containing 5 mg ortho-nitrophenylphosphate in 1.8 ml 0.01 M diethanolamine, 0.5 mM MgCl ₂ , pH 9.5) was added. After 30 minutes the wells were read at 405 nm with a Thermomax plate reader to quantitate bound biotinylated recombinant NIF-IFL.

Example 2

Isolation of Native Neutrophil Inhibitory Factor From

Hookworm Lysate

(A) Preparation of Hookworm Lysate

Frozen canine hookworms were obtained from Antibody Systems (Bedford, TX) . Hookworms were stored at -70°C until used for homogenate.

Hookworms were homogenized on ice* ^' in homogenization buffer [0.02M Tris-HCl pH 7.4, 0.05 M NaCl, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ , 1.0 x 10" ⁵ M dithiothreitol, 1.0 x 10 ^"5 M E-64 Protease Inhibitor (CAS 66701-25-5), 1.0 x

lO^M pepstatin A (isovaleryl-Val-Val-4-amino-3-hydroxy- 6-methyl-heptanoyl-Ala-4-amino-3-hydroxy-6-methyl- heptanoic acid, CAS 26305-03-3), 1.0 x 10" ⁵ M chymostatin (CAS 9076-44-2), 2.0 x 10" ⁵ M APMSF (amidinophenyl- methylsulfonyl fluoride-HCl) , 5% (v/v) glycerol] using a Tekmar Tissuemizer homogenizer. The protease inhibitors E64, pepstatin A, chymostatin, and APMSF were obtained from Calbiochem (La Jolla, CA) . Approximately 3-6 ml of homogenization buffer was used to homogenize each gram of frozen worms (approximately 500 worms) . Insoluble material was pelleted by two sequential centrifugation steps: 40,000 X g,, _^ at 4°C for 20 minutes followed by 105,000 x g _nuu -at 4°C for 40 minutes. The supernatant solution was clarified by passage through a 0.2 mm cellulose acetate filter (CoStar) .

(B) Concanavalin A Sepharose Chromatography of Hookworm

Lysate

Hookworm lysate (79 ml) was adsorbed to 16 ml of Concanavalin A Sepharose (Pharmacia) pre-equilibrated with Con A buffer [0.02 M Tris-HCl, pH 7.4, 1 M NaCl,

0.001 M CaCl ₂ , 0.001 M MnS0 ₄ , 1 x 10 ^"5 M dithiotreitol] by recycling it through a 1.6 x 8 cm column at a flow rate of 3 ml/min (90 cm/hour) for 2 hours. The column was at room temperature (24°C) while the reservoir of lysate was maintained on ice throughout the procedure. The column was subsequently washed with 80 ml of Con A buffer. The Con A buffer in the column was displaced with buffer containing 0.5 M methyl-alpha- mannopyranoside and flow stopped for 30 minutes. Flow was then restarted at a flow rate of 0.5 ml/min (15 cm/hour) . Material that had inhibitory. activity in neutrophil function assays was eluted "with approximately three column volumes of Con A buffer containing 0.5 M methyl-alpha-mannopyranoside (CAS 617-04-09) . The yield

of neutrophil adhesion inhibitory activity in this step was approximately 38%.

Figure 1 depicts Concanavalin A Sepharose chromatography of the hookworm lysate performed as described above. Absorbance at 280 nm was plotted as a function of time.

(C) Molecular Sieve Chromatography Using Superdex 200

Active fractions eluted from immobilized Concanavalin A (see step (B) above) and concentrated by ultrafiltration at 4°C using an Amicon stirred cell equipped with a 10,000 dalton cut-off membrane (YM10) , then 5-20 ml of the concentrate were loaded on a 2.6 cm x 60 cm column of Superdex 200 prep (Pharmacia) attached in series with an identical column (combined dimensions of 2.6 x 120 cm). Both columns were pre-equilibrated with 0.01 M potassium phosphate, pH 7.35, 0.150 M NaCl, 1 x 10" ⁵ M dithiotreitol at 24°C The chromatography was conducted at a flow rate of 1.5 ml/min; anti-adhesion activity typically eluted 395-410 ml into the run (K _1V of 0.46, see Fig. 2). This elution volume would be expected for a globular protein with a molecular mass of 50,000. The yield of neutrophil function inhibitory activity in this step was typically 70-80%. If the ionic strength of the chromatography buffer employed was decreased to 0.01 M sodium phosphate, pH 7.00 and 10% (v/v) glycerol added, the activity eluted substantially earlier (K _av = 0.34) suggesting that under such conditions the protein either aggregates or changes its conformation (assuming a larger Stoke's radius) . Figure 2 depicts Superdex 200 Chromatography of

Concanavalin A-Purified Hookworm Lysate. Absorbance at 280 nm is plotted versus elution volumes Active fractions eluted from immobilized Concanavalin A (see step (B) above) and concentrated by ultrafiltration at 4°C using an Amicon stirred cell equipped with a 10,000

dalton cut-off membrane (YM10) , then 5-20 ml of the concentrate were loaded on a 2.6 cm x 60 cm column of Superdex 200 prep (Pharmacia) attached in series with an identical column (combined dimensions of 2.6 x 120 cm). Both columns were pre-equilibrated with 0.01 M potassium phosphate, pH 7.35, 0.150 M NaCl, 1 x 10" ⁵ M dithiotreitol at 24°C The chromatography was conducted at a flow rate of 1.5 ml/min; activity eluted 395-410 ml into the run (K _ΪV of 0.46) .

(D) Ceramic-Hydroxyapatite Chromatography

Material purified by molecular sieve chromatography was concentrated five-fold by ultrafiltration using an Amicon stirred cell equipped with a 10 kilodalton cut-off membrane at 4°C and then diluted ten-fold with water. The desalted sample was loaded on a 0.8 x 10 cm column of ceramic hydroxyapatite ("HA") (Pentax, American International Chemical, Inc. , Natick, MA, 2 mm) equilibrated with 0.001 M potassium phosphate, pH 7.00, 1 x 10" ⁵ M CaCl ₂ , 1.0 x 10" ⁵ M dithiothreitol at 24°C The loading was conducted at a flow rate of 0.8 ml/min (95.5 cm/hour) . The column was developed with a 50 ml linear gradient of potassium phosphate ranging from 0.001 M to 0.0375 M at a flow rate of 0.5 ml/minute. Neutrophil inhibitory activity eluted sharply at 0.025 M potassium phosphate and then trailed to 0.0325 M potassium phosphate (fractions 37 to 48) . The yield of activity in this step was approximately 48%.

Figure 3 depicts Ceramic Hydroxylapatite Chromatography of Superdex/Concanavalin A-Purified Hookworm lysate plotting absorbance at 280 nm and potassium phosphate concentration versus fraction number. Neutrophil inhibitory activity eluted in fractions 37 to 48.

(E) Reverse Phase HPLC

Hookworm lysate fractionated by chromatography on Concanavalin A Sepharose, Superdex, and ceramic hydroxylapatite (-100 mg) was loaded on to a 0.48 x 15 cm column of 300 angstrom C4 (Vydac) which was then developed with a linear gradient of 0-60% acetonitrile in 0.1% trifluoroacetic acid at 1 ml/minute with a rate of 1% change in acetonitrile/minute. Neutrophil inhibitory activity typically elutes between 41 and 45% acetonitrile, the activity corresponding with a broad peak.

Figure 4 depicts the results of reverse phase HPLC of the Neutrophil Inhibitory Factor. Inhibitory activity eluted between 43 and 45% acetonitrile, the activity corresponding with a broad peak at 43-45 minutes.

Table I

Summary of Example Purification

Example 3

Isolation of the Neutrophil Inhibitory Factor From Hookworm Lysate Using Preparative Isoelectric Focusing

Hookworm lysate was partially fractionated and desalted by molecular sieve chromatography on a 2.6 cm x 60 cm column of Superdex 200 prep (Pharmacia) attached

I- m series with an identical column (combined dimensions of 2.6 x 120 cm). Both columns were pre-equilibrated with 0.01 M sodium phosphate, pH 7.00, 10% (v/v) glycerol at 24°C Adhesion inhibiting fractions eluting

at 350-370 ml were diluted to 55 ml by the addition of 1.4 ml of 40% Biolyte 3-10 ampholyte (BioRad) and 10% (v/v) glycerol. This mixture was focused with a constant power of 12 W for 5 hours at 4°C in a Rotofor preparative isoelectric focusing prep cell (BioRad) .

Twenty fractions were harvested; inhibitory activity was detected in fractions 6-9, corresponding to an isoelectric point of 4.5. The overall yield of inhibitory activity for this step was approximately 30%.

Example 4

Ion Exchange Chromatography

Hookworm lysate fractionated by molecular sieve chromatography on Superdex 75 (Pharmacia) was mixed with an equal volume of Mono Q buffer [0.02 M Tris-HCl, pH 7.5] and loaded on to a 0.5 x 5.0 cm Mono Q anion exchange column (Pharmacia) equilibrated with Mono Q buffer at a flow rate of 1 ml/minute (306 cm/hour) . The column was then developed with a linear gradient of 0-0.5 M NaCl in column buffer at 0.5 ml/minute (153 cm/hour) . Neutrophil inhibitory activity consistently eluted at 0.4 M NaCl. The overall yield of inhibitory activity for this isolation was about 2-5%.

Example 5

SDS-Polvacrylamide Gel Electrophoresis The protein composition of hookworm lysate and fractionated lysate was analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli, U.K. 1970, Nature 227, 680) after silver staining (Morrisey, J.H. 1981, Anal. Biochem. 117, 307) . Samples were mixed with an equal volume of 20% glycerol,

5% SDS, and 0.125 M Tris-HCl, pH 6.8 »ήd placed in a boiling water bath for 5 minutes. Samples were subsequently applied onto 10% SDS polyacrylamide slab

gels of 0.75 mm thickness and subjected to electrophoresis for 2 hours at constant voltage (125 V) . Figure 5 depicts the results of SDS polyacrylamide gel electrophoresis. Samples were applied to a 10% polyacrylamide slab gel (Novex, La Jolla, CA) . Lanes 1-10, left to right, are (1) molecular weight standards; (2) molecular weight standards; (3) HPLC pool of HA fractions #37-41, non-reduced; (4) blank; (5) HPLC pool of HA fractions #37-41, reduced; (6) blank, (7) HPLC pool of HA fractions #37-41, reduced, (8) HPLC pool of HA fractions #37-41, non-reduced; (9) HPLC pool of HA trailing fractions #42-48, non-reduced, (lθ)molecular weight standards. The molecular weight standards used were: myosin, 200,000 (rabbit muscle); beta-galactosidase, 116,300 (E. coli) ; phosphorylase b, 97,400 (rabbit muscle); bovine serum albumin, 66,300; glutamic dehydrogenase, 55,400, (bovine liver); carbonic anhydrase, 31,000, (bovine erythrocyte) ; trypsin inhibitor, 21,500, (soybean). Following the last step of the isolation procedure

(reverse phase HPLC) only a single diffuse band with an apparent molecular weight ranging from 33,000 to 47,000 was observed upon SDS-PAGE (see Fig. 5) . When 50 mM dithiothreitol was added to the sample prior to boiling, the diffuse band migrated with an estimated molecular weight of 43,000 to 54,000.

Example 6

Laser-Desorption Time-of-Flight Mass Spectrometry of the

Isolated Neutrophil Inhibitory Factor The estimated mass for the NIF isolated as described in Example 2(E) was determined using laser-desorption time-of-flight mass sμectrometry.

A 1 ml aliquot of the sample was diluted with an equal volume of a saturated solution of 3,5-dimethozy-4-hydroxy-cinnamic acid dissolved in 30%

agueous CH ₃ CN, 0.1% TFA. The diluted sample was spotted onto a copper sample stage and allowed to air dry. Mass analysis was performed using a Shimadzu LAMS-50KS laser desorption time of flight mass spectrometer (Shimadzu Corp. , Kyoto, Japan) . Ionization of the sample was accomplished by focusing 500 laser pulses (355 nm, pulse width <5 nsec) from a Nd-YAG laser (Spectra-Physics, Inc., Mt. View, CA) onto the sample stage. The resulting ions were accelerated into the mass spectrometer by a 5 kV potential. Calibration of the instrument was accomplished using standard proteins of known mass.

Figure 6 depicts the results of laser-desorption time-of-flight mass spectrometry of the isolated neutrophil adhesion inhibitor. Five picomoles of the purified neutrophil function inhibitor was analyzed with a laser desorption time-of-flight mass spectrometer. The estimated mass was determined as 41,200. A small fraction of the sample had a mass of 82,400; this was interpreted to be a dimer.

Example 7

Neutrophil Inhibitory Factor is a Glycoprotein

Purified NIF (prepared according to Example 2(E)) (-2 mg) was electrophoresed in a 10% SDS polyacrylamide gel and the resolved protein transferred by Western blotting (Towbin, et al., 1979 Proc. Natl. Acad. Sci. (USA) 26, 4350-4354) to a Zeta-Probe ^® nitrocellulose membrane (BioRad, Emeryville, CA) . The membrane was treated as described in the instructions to the GlycoTrack™ Kit (Oxford GlycoSystems, Rosedale, NY) to oxidize carbohydrates to aldehydes which were then reacted with biotin-hydrazide leading to incorporation of biotin into any carbohydrate present. Biotinylated carbohydrate was subsequently detected by reaction with a streptavidin-alkaline phosphatase conjugate.

Visualization was achieved using a substrate which reacts with alkaline phosphatase bound to glycoproteins on the membrane, forming a colored precipitate. Neutrophil Inhibitory Factor was stained using this method, demonstrating that it contained carbohydrate and is therefore a glycoprotein.

Example 8

Organic Extraction of the Hookworm Lysate

One milliliter of hookworm homogenate known to have inhibitory activity in the neutrophil-plastic adhesion assay was extracted by vortexing 1 minute with 1 ml of a chloroform/methanol (2:1) mixture in a 15 ml glass Corex test tube. The organic layer was removed and dried under a stream of nitrogen gas. Residual lipids were resuspended in 0.5 ml HSA assay buffer by sonication for

2 minutes (Branson Model 1200, Danbury, CT) . Resuspended lipids had no inhibitory activity in the neutrophil-plastic adhesion assay when tested at a final dilution of 1:2.

Example 9

Production And Determination Of The Amino Acid Sequence Of Peptide Fragments Of Neutrophil Inhibitory Factor

Samples of NIF were obtained as described in Example 2. Two separate volumes, each containing approximately 10 mg NIF, were first degassed on a Speed

Vac until the samples were frozen and then lyophilized. The dried samples were resuspended in 50 mM N-ethylmorpholine, pH 8.5, and digested with either endoproteinase AspN (Boehringer Mannheim, Indianapolis, IN) , Lys C (Boehringer Mannheim, Indianapolis, IN) or trypsin (Worthington, Freehold, NJ) at af substrate to enzyme ratio of 25:1. Incubation was at ambient temperature for 24 hours and a small amount of isopropanol was added to the digestion mix to prevent

microbial contamination. At the end of the digestion, the samples were degassed on a Speed Vac and dried by lyophilizing. The digestion mixtures were resuspended in 6M guanidine/HCl for fractionation of peptides by reversed phase HPLC (RP HPLC) . Peptides were isolated by RP HPLC on a ToyoSoda 120T C18 (4.5 X 250 mm) column using an LKB HPLC system with Kratos (ABI, Foster City, CA) detectors. The column was developed with a linear gradient of acetonitrile in 0.1% trifluoroacetic acid (TFA) . The gradient was from 5 to 54% acetonitrile over 120 minutes at a flow rate of 0.5 ml/minute. Peptide peaks monitored by A ₂₀₆ and A^g were collected using an LKB SuperRac with calibrated peak detection. The collected fractions were neutralized with ammonium carbonate, 20 mg SDS was added, and the fractions dried under N ₂ before sequencing. Peptides were sequenced on a 470A/120A/900A gas phase sequencer (ABI, Foster City, CA) . Residue identification was performed manually by analysis of the HPLC chromatograms and quantification of the PTH residues was performed by online analysis on the

900A computer. Cysteine residues were not detected in this analysis because the protein had not been alkylated. In experiments in which the protein was digested with trypsin, the protein was alkylated with vinylpyridine before fragmentation, thereby permitting the detection of cysteine in the tryptic fragments. Aspartic acid and tryptophan residues were identified but not quantitated because background peaks overlapped the PTH residues in the HPLC elution. The initial yields ranged from 1 pmole to 10 pmole and the repetitive yield was usually between 92 and 95%. Figure 7 depicts the amino acid sequences that were obtained from the proteolytic fragments. In Fig-tire 7, positions enclosed in parentheses were not determined with absolute certainty. Abbreviations for amino acids beginning with a capital letter were observed in higher

yield and are preferred in these cases. The abbreviation Xxx indicates an undetermined amino acid at that position, since no specific amino acid was identified during Edman degradation of the peptide. See Scarborough et al. J. Biol. Chem 266:9359. 1991.; Perin et al., J. Biol. Chem. 266:3877. 1991.

Example 10

Cloning and Sequencing of Neutrophil Inhibitory Factor from Hookworm NIF was cloned from a canine hookworm cDNA library, constructed as follows: Total RNA was isolated from whole hookworms by guanidium thiocyanate extraction (McDonald et al. , Meth. Enzymol. 152:219 (1987)). Poly(A)+ RNA was purified from 500 mg of total-hookworm RNA using oligo d(T) cellulose affinity chromatography (PolyA Quik; Stratagene, La Jolla, CA) . Double stranded cDNA was synthesized from poly(A)+ RNA using random hexa er primers and avian myoblastosis virus (AMV) reverse transcriptase (Amersham, Arlington Hills, IL) . cDNA fragments larger than 1 kilobase pairs were purified on a 6% polyacrylamide gel and ligated to EcoRI linkers (Stratagene) using standard procedures. Linkered cDNA was ligated into lambda gtlO (Stratagene, La Jolla, CA) and packaged using Gigapack Gold II (Stratagene) .

Double stranded cDNA probes for hookworm NIF were generated by polymerase chain reaction from hookworm RNA using primers derived from NIF peptide sequences. The sequences obtained for two NIF peptides (see Fig. 7) , T-20 (Leu-Ala-Ile-Leu-Gly-Trp-Ala-Arg) and T-22-10

(Leu-Phe-Asp-Arg-Phe-Pro-Glu-Lys) , were used to design primers 30.2 and 43.3.RC, respectively. <*The sequences of 30.2 and 43.3.RC were

5'-CTCGAATTCT(GATC)GC(ATC)AT(ATC) (CT)T(GATC)-GG(ATC)TGGG C-3' and

5'-CTCGAATTCTT(TC)TCTGG(GA)AA-(GA)CG(GA)TC(GA)AA-3', respectively. Bracketed positions represent redundant nucleotides. Single stranded cDNA was synthesized by priming 1 mg of hookworm poly(A)+ RNA (preparation described above) with random hexanucleotides and extending with AMV reverse transcriptase (Amersham, Arlington Hills, IL) . One twentieth of the reaction product was amplified using the PCR GeneAmp kit (Perkin Elmer, Norwalk, CT) , with 400 pmol of each of 30.1 and 43.RC (manufactured by Research Genetics, Huntsville, AL) , on a Perkin Elmer DNA Thermal Cycler. PCR conditions were: cycles 1-2, denaturation at 94°C for 2 minutes, annealing at 58°C for 2 minutes and elongation at 72°C for 2 minutes; cycles 3-42, denaturation at 94°C for 45 seconds, annealing at 58°C for 45 seconds and elongation at 72°C for 2 minutes. The -430 base pair amplification product, referred to as the 30.2/43.3.RC fragment, was separated from reaction contaminants by electroelution from a 6% polyacrylamide gel (Novex, San Diego, CA) . The 30.2/43.3.RC fragment was labelled with [α- ³² P]-dCTP (Amersham) using random primer labelling (Stratagene, La Jolla, CA) ; labelled DNA was separated from unincorporated nucleotides using a ChromaSpin-lO column (Clontech, Palo Alto, CA) . Prehybridization and hybridization conditions were

6X SSC (SSC: 150 mM NaCl, 15 mM trisodium citrate) , 0.02 M sodium phosphate pH 6.5, 5X Denhardt's solution, 0.5% (w/v) SDS, 0.01 M EDTA, 100 mg/ml sheared, denatured salmon sperm DNA, 0.23% dextran sulfate, 50% formamide. Prehybridization and hybridization were at

42°C, and the filters were washed for 20 minutes with 0.2X SSC at 60°C after two prewashes with 2X SSC for 15 minutes. The filters were exposed overnight to X-ray film with two intensifying screens at -70°C Approximately 300,000 recombinant phage of the random primed hookworm library (unamplified) were

screened with the 30.2/43.3.RC NIF PCR fragment. About 120 recombinant phage hybridized to this probe, of which seven were isolated for nucleotide sequencing analysis. Double stranded sequencing was effected by subcloning the EcoRI cDNA fragments contained in these phage isolates into pBluescript II vector (Stratagene, La Jolla, CA) . DNA was sequenced using the Sequenase version 2.0 kit (U.S. Biochemical, Cleveland, OH) and synthetic oligonucleotide primers. The NIF phage isolates contained DNA that encoded polypeptides that bore striking resemblance to the amino acid sequences obtained for purified NIF (see Figure 7) . Figure 8 depicts the nucleotide sequence of the coding region of Neutrophil Inhibitory Factor cDNA (clone IFL) and its predicted amino acid sequence. A single isolate, NIF-IFL, encoded an open reading frame of 825 nt, initiating with a methionine and terminating with a TGA stop codon (Figure 8) . The NIF polypeptide encoded by NIF-IFL is 274 amino acid residues with a calculated molecular weight of 30,680 daltons. Figure 9 depicts the alignment of the predicted amino acid sequences of several Neutrophil Inhibitory Factor isoform clones. Each line of sequence represents the corresponding sequence segments of the various clones isolated. Each segment is identified by its clone designation (e.g.,

IFL, 3P, 2FL, 3FL, 4FL, 6FL and IP) . The complete amino acid seguence of clone IFL is listed in standard three-letter amino acid code at the top of each sequence segment. Clones having the same amino acid in a given position as clone IFL are denoted by ".". Amino acid substitutions are indicated by the appropriate three-letter code. " " indicates a space inserted to maintain alignment of the sequences. Tbe carboxy termini of the IFL and IP sequences are denoted by an asterisk. The other six NIF phage isolates encoded partial NIF polypeptides; that is they did not contain

either an N-terminal methionine residue or a C-terminal stop codon, as compared to the NIF-IFL polypeptide (Figure 9) . These partial NIF isolates comprised six predicted NIF isoforms that were significantly similar to, but not identical to the prototypical NIF-IFL polypeptide.

Example 11

Expression of Functional Recombinant Neutrophil

Inhibitory Factor bv Mammalian Cells (A) Transient Expression in COS-7 Cells.

The segment of DNA encoding the NIF-IFL isoform was amplified from the original λgtlO isolate DNA using unique primers for the 5'- and 3'-ends of the coding region. The 5'-primer was composed of a restriction endonuclease site (EcoRl) , a consensus ribosome binding site (Kozak, M. , Cell 44.: 283 (1986)), the ATG initiation codon of NIF and the succeeding 6 codons of the gene. The 3'-primer was composed of a unique nucleotide sequence to the 3'-side of the TGA termination codon of NIF and a restriction endonuclease site (EcoRl) . The nucleotide sequences of the 5'- and 3'-primers were

5'-ACC-GAA-TTC-ACC-ATG-GAG-GCC-TAT-CTT-GTG-GTC and 5'-CTG-GAA-TTC-TCG-CTT-ACG-TTG-CCT-TGG-C, respectively. Five microliters of the lambda plaque suspended in 1 ml dilution buffer were used as template DNA. Amplification was accomplished using the PCR GeneAmp kit (Perkin Elmer, Norwalk, CT) , with 400 pmol of each of the 5'- and 3'-primers (manufactured by Research

Genetics) , on a Perkin Elmer DNA Thermal Cycler. The PCR conditions were: cycle 1, denaturation at 97°C for 1 minute, primer annealing for 1 minute at 37°C, ramp from 37°C to 72°C in 2 minutes, and amplification for 2 minutes at 72°C; cycles 3 and 4, denaturation at 94°C

for 1 minute, primer annealing for 1 minute at 37°C, ramp from 37°C to 72°C in 2 minutes, and amplification for 2 minutes at 72°C; cycles 5 through 34, denaturation at 94°C for 1 minute, primer annealing for 1 minute at 45°C, and amplification for 2 minutes at 72°C.

The amplification product (887 bp) was separated from reaction contaminants using a ChromaSpin 400 column (Clontech Laboratories, Inc. Palo Alto , CA) . The ends of the amplification product were trimmed with the restriction endonuclease EcoRl and the resulting fragment of DNA (875 bp) ligated into EcoRl-digested plasmid pSG5 (Stratagene, La Jolla, CA) using standard techniques. The resulting ligation mixture was used to transform SURE!™ competent cells (Stratagene, .La Jolla, CA) .

An isolate containing the 875 bp insert in the proper orientation (5'-end of the coding region proximal to the pSG5 SV40 promoter) was grown in 250 ml Circle Grow™ (Biolo, San Diego, CA) with 50 mg/ml ampicillin and plasmid DNA was prepared using a Magic Maxi Prep™ DNA purification system (Promega, Madison, WI) . Ten micrograms of purified plasmid DNA was transferred into 3.5 x 10° COS7 cells (ATCC No. CRL 1651) by electroporation (0.4 cm electroporation cell, 325 V, 250 F, infinite resistance, 0.5 ml cells at 7 x 10 ⁶ /ml in Hepes buffered saline, pH 7.0, 4°C) . After electroporation the cells were allowed to stand on ice for 2 to 3 minutes before dilution with 14 ml warm DMEM:RPMI 1640 (1 to 1 ratio) supplemented with 10% fetal bovine serum prewarmed to 37°C The cells were placed in 100 mm cell culture dishes and incubated at 37°C with 8% C0 ₂ . Cell culture supernatant fluid was removed at 1, 2 and 3 days after plating and assayed for NIF activity.

(B) Detection and Quantitation of Neutrophil Inhibitory Factor Activity in Cell Culture Medium.

15 ml of cell culture fluid was harvested from electroporated C0S7 cells (pSG5/NIFlFLCRl) . When assayed directly using the neutrophil-plastic adhesion assay (Example 1(C)), this fluid exhibited neutrophil inhibitory activity to dilutions as great as 1:8. An IC _j0 at approximately 1:14 was determined using the hydrogen peroxide release assay (Example 1(E)). No activity was observed using cell culture fluid harvested from COS7 cells electroporated with a control expression plasmid (pCAT; Promega, Madison, WI) .

(C) Stable Expression in CHO Cells (1) Preparation of Plasmid DNAs The NIF-IFL insert in the pSG5 construct described above in section (A) was excised by digestion with the restriction endonuclease EcoRl. The 875 bp NIF-IFL fragment was gel purified (Magic PC Prep, Promega, Madison, WI) and ligated into EcoRl digested pBluescript II KS (Stratagene, La Jolla, CA) using standard techniques. The resulting ligation mixture was used to transform SURE™ competent cells as described by the supplier (Stratagene, La Jolla, CA) . Transformed cells were plated on LB agar containing IPTG and X-gal (Sambrook, Fritsch, and Maniatis, Molecular Cloning, a Laboratory Manual, Cold Spring Harbor Laboratory Press, 1989, pp. 1.85 to 1.86). White colonies were screened for plasmids containing the NIF-IFL insert in the proper orientation by digesting plasmid DNA with BamHI; those colonies harboring a plasmid that yielded a 200 bp BamHI fragment were retained. Plasmid was prepared from one of these colonies (Magic Maxi Prep, ProMega, Madison, WI) and digested with Hindlll and Notl to yield a NIF-IFL fragment with a Hindlll overlap on the 5'-end and a Notl overlap on the 3'-end. This DNA fragment was

gel purified and ligated into Hindlll-NotI digested pRC/CMV (Invitrogen, San Diego, CA) . The resulting ligation mixture was used to transform SURE™ competent cells. Milligram quantities of pRC/CMV-NIF-lFL were prepared using the Magic Maxi Prep kit.

The plasmid pLTRdHFR26 (Mol. Cell. Biol. 3:32-43 (1983), Nature 275:617-623 (1978)) in the E. coli strain RRI was purchased from the ATCC (American Type Culture Collection, Rockville, MD) . Plasmid DNA was purified from a chloramphenicol amplified culture (Sambrook,

Fritsch, and Maniatis, Molecular Cloning, a Laboratory Manual, Cold Spring Harbor Laboratory Press, 1989, p. 1.33) using the Magic Maxi Prep kit.

(2) Transfection of CHO Cells Chinese hamster ovary (CHO) cells harboring a defect in the dihydrofolate reductase gene (dhfr) were obtained from the ATCC (catalog number: CRL 9096) and grown in a 1:1 mixture of RPMI 1640 and DMEM media (Irvine Scientific, Santa Ana, CA) supplemented with 10% FBS, 10 mM HEPES, essential amino acids, nonessential amino acids, 5 x 10' ⁵ M j8-mercaptoethanol, 10 mM sodium pyruvate, and 2 mM glutamine (combo medium) . The CHO cells were transfected using the Calcium Phosphate Transfection System following the manufacturer's instructions (Gibco BRL, Gaithersburg, MD) in 10 cm cell culture dishes at the following ratios: 1 x 10 ⁶ cells, 20 g pRC/CMV-NIFlFL DNA and 5 g pLTRdHFR26 DNA. The cells were incubated for 12 hr at 37'C in the presence of the co-precipitated DNAs.

(3) Selection and Amplification of CHO-NIFIFL Clones The adhered cells were washed once with fresh combo medium and placed under combo medium containing 500 g/ml of the antibiotic G418 (Gibco BRL, Gaithersburg, MD) for

two weeks at 37 ^* C to select for those clones exhibiting neomycin resistance. The G418 resistant cells were trypsinized from the culture dishes using standard techniques and washed once with fresh combo medium containing 500 g/ml G418. The washed cells were allowed to attach to 10 cm cell culture dishes (1 x 10 ⁶ per dish) and covered with combo medium containing 500 g/ml G418 and 10, 20, 40, 60, or 80 nM methotrexate (Sigma, St. Louis, MO) . After two weeks incubation at 37°C, the dishes were examined for colonies and the culture supernatant fluids assayed for NIF activity using the calcein assay for neutrophil adhesion. The cells in dishes exhibiting NIF activity were trypsinized and washed as before and plated to obtain single colonies. One single cell isolate, designated 8F5, expressing NIF activity was chosen for further methotrexate amplification.

8F5 cells were grown to confluence in a 10 cm cell culture dish, detached with trypsin and washed as before. The cells were diluted with combo medium containing 500 g/ml G418 and 1 x 10° cells were placed in 10 cm culture dishes. The adhered cells were covered with combo medium containing 500 g/ml G418 and 40, 80, 160, 320, or 640 nM methotrexate. Again, after two weeks incubation at 37"C, the dishes were examined for colonies and supernatant fluids assayed for NIF activity. Cells were released from the plates by trypsin treatment as before. The pool of cells obtained from the 320 nM methotrexate dish was selected for further use and single cell isolates were obtained.

Three successive rounds of single cell isolation were performed on each colony to ensure clonal, purity. The final cell line, designated , produces NIF at greater than 50 g/ml in the presence of 500 μg/ml G418 and 320 nM methotrexate.

(D) Fractionation of Neutrophil Inhibitory Factor Activity bv Chromatography on Immobilized Concanavalin A.

Five ml of C0S7(pSG5/NIFlFLCRl) cell culture fluid was mixed with an equal volume 0.02 M bis

Tris-propane-HCl, pH 7.3, 1 M NaCl, 0.001 M CaCl ₂ , 0.001 M MnS0 ₄ and loaded onto a one ml column of Concanavalin A Sepharose (Pharmacia, Piscataway, NJ) equilibrated with the same buffer. The sample was cycled through the column in a closed loop for 1 hour at 2 ml/minute at 20°C. The column was subsequently washed with 5 ml of 0.02 M bis Tris-propane-HCl, pH 7.3, 1 M NaCl, 0.001 M CaCl ₂ , 0.001 M MnS0 ₄ . The buffer resident in the column was displaced with buffer containing 0.5 M methyl-alpha-mannopyranoside and flow stopped- for 15 minutes. Flow was restarted at 1 ml/minute and approximately 11 ml of sugar-containing eluate collected. The eluate was dialyzed 18 hours against 1 liter 10 mM potassium phosphate, pH 7.35, 150 mM NaCl at 4°C and concentrated to 1.1 ml using an Amicon centrifugal concentrator equipped with a 10,000 molecular weight cut-off membrane (CentriPrep 10, Amicon, Beverly, MA) . When assayed by the neutrophil-plastic adhesion assay (Example 1(C)), this sample exhibited substantial activity at a dilution of

1:16, indicating that a significant portion of the neutrophil function inhibitor activity present in the cell culture fluid binds to immobilized Concanavalin A. This behavior is identical to that observed for crude extracts of Ancylostoma caninum (Example 2(B)) and is consistent with the inhibition resulting from the synthesis and secretion from transfected mammalian C0S7 cells of a glycoprotein that acts as an^ ^' inhibitor of neutrophil function. As a control, 5 ml of C0S7 cell culture medium from cells electroporated in the absence of DNA was

chromatographed on Concanavalin A Sepharose in the same manner as described above. No activity was observed after Concanavalin A-Sepharose chromatography using the neutrophil-plastic adhesion assay (Example 1(C)).

(E) Fractionation of Neutrophil Inhibitory Factor

Activity bv Anion Exchange Chromatography using POROS II 0/M.

Five ml of C0S7(pSG5/NIFlFLCRl) cell culture fluid was dialyzed 18 hours against one liter of 10 mM bis Tris-propane-HCl, pH 7.0 at 4°C and loaded at 3 ml/minute onto a 0.46 x 10 cm column of Poros II Q/M (PerSeptive Biosystems, Inc., League City, TX) equilibrated with the same buffer. The column was washed with one column volume of equilibration buffer and developed with a linear gradient of sodium chloride from 0 to 0.5 M over 14.4 column volumes collecting 2 ml fractions. Significant activity in the neutrophil-plastic adhesion assay (Example 1(C) was detected in fractions 17 and 18, corresponding to about 0.45 M NaCl. When fractions were concentrated twenty-fold using centrifugal concentrators equipped with a 10,000 MWCO membrane (Amicon MicroCon 10, Beverly, MA) , substantial activity was found in fractions 16-19. Neutrophil inhibitory factor present in extracts from Ancylostoma caninum elutes likewise from an anion exchange column (Mono Q, Pharmacia, Piscataway NJ) at 0.4 M NaCl (Example 4).

Example 12 Expression of Functional Recombinant Neutrophil

Inhibitory Factor in Pichia pastoris (A) Description of the Pichia shuttle/expression vector.

The Pichia strain GTS115 (his4) (Stroman, D.W. et al., U.S. Patent No. 4,655,231 (August δ, 1989)) and the E. coli-Pichia shuttle vectors pHILSl and pHILD5 referred to hereafter are part of the Pichia yeast expression system licensed from the Phillips Petroleum Company (Bartlesville, Oklahoma) .

All of the Pichia manipulations were performed essentially as described for Saccharo yces cerevesiae in Gene Expression Technology, pp.231-471, Academic Press, New York, (D.V. Goeddel, edit. 1991) and in Stroman, D.W. et al., US Patent No. 4,855,231 (August 8, 1989) . The pHIL7SP8 vector used to direct expression of NIF in P. pastoris was assembled from pHILSl and pHILD5 and from synthetically generated fragments. The pHIL7SP8 plasmid contained the following elements cloned onto pBR322 sequences:

1) 5' AOXl, about 1000 bp segment of the P. pastoris alcohol oxidase 5' untranslated and promoter sequences (see Stroman, D.W. et al., U.S. Patent No. 4,855,231 (August 8, 19δ9) the disclosure of which is incorporated herein by reference) .

2) The PH01 P. pastoris secretion signal.

3) A 19-amino acid synthetic pro-sequence fused to the PHOl signal. This pro-sequence represents one of the two 19-aa pro-sequences designed by Clements et al.,(1991. Gene, 106:267-272) on the basis of the yeast alpha-factor leader sequence.

4) A synthetic multi-cloning site.

5) 3' AOXl, about 256 bp segment of the aoxl terminating sequence (see Stroman, D.W. et al., U.S.

Patent No. 4,655,231 (August δ, 1969) the disclosure of which is incorporated herein by reference) .

6) P. pastoris histidinol dehydrβgenase gene, hiε4 , contained on a 2.4 kb fragment to complement the defective his4 gene in the host GTS115 (see Stroman,

D.W. et al., U.S. Patent No. 4,855,231 (August 8, 1989)

the disclosure of which is incorporated herein by reference) .

7) Region of 3' AOXl untranslated DNA sequence, which together with the 5' AOXl region is necessary for site-directed integration (see Stroman, D.W. et al., U.S. Patent No. 4,855,231 (August 8, 1969) the disclosure of which is incorporated herein by reference) .

(B) Construction of PHIL7SP-NIC1/PHIL7SP-NIC10 and expression in Pichia.

The segment of DNA encoding NIF was PCR-amplified from a sub-clone of NIF-IFL in Bluescriptll (Stratagene, La Jolla, CA) using unique primers for the 5'- and 3'-ends of the coding region. The 5'-primer contained no restriction endonuclease sites and corresponded to the region beginning at the 5'-end of proteolytically processed NIF and the succeeding 7 codons. The codon for the first residue of the mature NIF was altered from AAT to AAC (both codons translate to asparagine). The 3'-primer was composed of 8 codons at the 3' end of the coding region, a TAA stop replacing the TGA stop of the natural gene, and three unique restriction endonuclease sites (Hindlll. Spel, and Bglll) . The sequences of the 5'- and 3'-primers used were 5'-AAC-GAA-CAC-AAC-CTG-AGG-TGC-CCG and

5'-CCT-CCT-CCT-AGA-TCT-AAG-CTT-ACT-AGT-TTA-TAA-CTC-TCG-G AA-TCG-ATA-AAA-CTC, respectively.

Amplification was accomplished using 100 pmol of each primer, 2 units of Vent polymerase in IX Vent buffer (New England Biolabs, Beverly, MA), and 0.2 mM of each of dATP, dCTP, dGTP, and dTTP. One hundred nanograms of Bluescriptll-containing NIF-IFL were used as template DNA. The PCR conditions were the same for all ten cycles: denaturation at 95°C for 1 minute, primer annealing at 60°C for 1 minute, and amplification

for 1.5 minutes at 72°C. The amplification product was purified as described above and digested with Bglll. The amplification product was then ligated into Stul-BgJ.il cleaved pHIL7SP8 using standard methods. The ligation mixture was used to transform E.coli WK6, and ampicillin resistant clones were obtained on ampicillin plates. Based on restriction and DNA sequence analysis, correct insert sequences in two of the resulting plasmid clones, pHIL7SP-NIlcl and pHIL7SP-NIlcl0, were selected to transform the P.pastoris yeast strain GTS115 (his4) . These vectors were digested with either Notl (targeting integration to the expression cassette in the AOXl region) or Sail (targeting integration to the HIS4 locus) . The 4 restricted DNA preparations were introduced individually into Pichia by electroporation, essentially as described by Becker, D. and Guarente, L. , Methods in Enzymology, vol. 194, pp. 182-169 (1991). Briefly, the cells were grown in YEPD medium at 30°C to an 00^ of 1.3 to 1.5. The cells were pelleted at 4°C (1500 x g for 5 minutes) and resuspended in 500 ml ice cold sterile distilled water. The cells were pelleted as above and resuspended in 250 ml ice cold distilled water. After the cells were pelleted again, they were resuspended in 20 ml ice cold 1 M sorbitol. After a final pelleting the cells were resuspended in 1 ml ice cold 1 M sorbitol. Forty μl cells in 1 M sorbitol were mixed with 5 μl of linearized DNA and the mixture transferred to an ice cold 0.2 cm gap electroporation cuvette. After 5 minutes on ice, the cells were pulsed at 50 uF,

1.5 kV/c , and 200 Ω resistance. One ml of ice cold 1 M sorbitol was added to the cuvettes and 100 to 500 ul of the cell suspension were spread on miniβial dextrose plates. The plates were incubated at 30°C until colonies appeared. The transformation mix was plated on minimal dextrose (MD) medium to select for His+

transformants. Subsequent selection for NIF expression was performed in shake flask cultures in minimal medium containing methanol as described in Stroman, D.W. et al., U.S. Patent No. 4,855,231 (August 8, 1989)

(C) Detection and Quantitation of Neutrophil Inhibitory

Activity in Cell Medium.

Pichia cell supernatant (pHIL7SP-NlclO) was obtained by centrifugation for 15 minutes at 1,800 x g _max from cells 48 hours following methanol induction and filtered through a 0.22 μm cellulose acetate membrane. The filtered cell supernatant solution was concentrated about 3-fold using centrifugal concentrators equipped with a 10,000 MWCO membrane (Amicon MicroCon 10, Beverly, MA) and desalted by gel filtration using a 1 x 10 cm column of G-25 Sephadex Superfine (Pharmacia,

Piscataway, NJ) . Using the neutrophil-plastic adhesion assay (Example 1(C)), the desalted supernatant solution (diluted 2x by gel filtration) exhibited neutrophil inhibitory activity to dilutions as great as 1:640. No activity was observed using cell supernatant solution similarly harvested and treated from Pichia cells expressing a recombinant anti-thrombotic protein devoid of neutrophil inhibitory activity.

(D) Purification of Neutrophil Inhibitory Factor from Pichia

Following methanol induction for 46 hours, 75 ml of Pichia cell supernatant (pHIL7SP-NlclO) 48 hours following methanol induction was obtained by centrifugation for 15 minutes at 1,800 x g,, _^ and filtered through a 0.22 μm cellulose acetate membrane.

This was concentrated using an Amicon stirred UF cell equipped with a 10,000 molecular weight cut-off membrane (YM10) and then diluted with water (about 10-fold) . This diafiltration process was repeated until the

conductivity was reduced from 45 mS to 1 mS. The final volume of the concentrate was 25 ml.

This concentrate was dialyzed at 4°C for 6 hours against one liter of 0.05 M bis Tris-propane-HCl, pH 7.0 to adjust the pH to neutrality, and then against two changes of one liter of 0.001 M potassium phosphate, pH 7.0.

Fifteen ml of the dialyzed cell supernatant was loaded onto a O.δ x 15 cm column of ceramic hydroxyapatite (Pentax, 2 μm; American International Chemical, Inc., Natick, MA) equilibrated with 0.001 M potassium phosphate, pH 7.0 at a flow rate of 0.4 ml/min (48 cm/hour) . The column was washed with one column volume of 0.001 M potassium phosphate, pH 7.0 and then developed with a linear gradient from 0.001 tβ 0.050 M potassium phosphate over 20 column volumes at a flow rate of 0.35 ml/min. Substantial neutrophil inhibitory activity eluted at approximately 0.02 - 0.035 M potassium phosphate in much the same fashion as observed for neutrophil inhibitory factor isolated from Ancylostoma caninum (Example 2(D) ) .

Fractions exhibiting substantial neutrophil inhibitory activity (assessed using the neutrophil-plastic adhesion assay (Example 1(C))) were combined and concentrated to about 3 ml using an Amicon centrifugal concentrator equipped with a 10,000 molecular weight cut-off membrane (CentriPrep 10, Amicon, Beverly, MA) and applied to a 1 x 25 cm C4 300 A reverse phase column (5 μm particle size, Vydac, Hesperia, CA) equilibrated with 0.1% trifluoroacetic acid. The column was washed with four column volumes of equilibration buffer and then developed with a linear gradient of acetonitrile from 15 to 40%^όver 10 column volumes at a flow rate of 5 ml/min. A major complex peak absorbing at 214, 254, and 280 nm eluted at about 36-38% acetonitrile.

Fractions including and bracketing this peak were dried using a centrifugal evaporator to remove solvent and trifluoroacetic acid and rehydrated with 0.065 M potassium phosphate, pH 7.0, 0.08 M NaCl. The rehydrated fractions possessed substantial neutrophil inhibitory activity as judged by the neutrophil-plastic adhesion assay (Example 1(C)) and the hydrogen peroxide release assay (Example 1(E)).

Fractions with substantial activity were combined and sequenced by Edman degradation using a

470A/120A/900A gas phase sequencer (ABI, Foster City, CA) (See Example 9) and yielded the following sequence:

Asn-Glu-His-Asn-Leu-Arg-Xxx-Pro-Gln-Xxx-Gly-Thr- Glu-Met-Pro-Gly-Phe-Xxx-Asp-Ser-Ile-Arg-Leu-Gln-Phe-Leu- Ala-Met-His-Asn-Gly-Tyr-Arg-Ser-Lys-Leu-Ala-Leu- Gly-His-Ile-Se r-Ile-Thr-Glu. "Xxx" refers to an undetermined amino acid at that position, since no specific amino acid was identified during Edman degradation of the peptide. This sequence matches the predicted N-terminal sequence of NIF-IFL, the NIF isoform used in this construction construct (pHIL7SP-NlclO; see Figure 8) . The first position at which a residue was not detected is predicted to be a cysteine; cysteine residues could not be detected in this analysis because the protein had not been alkylated. The two other positions at which residues were not detected correspond to asparagine residues followed by either a serine or threonine one residue distant. This is a glycosylation consensus sequence [Asn-Xxx-(Ser/Thr) ] and the fact that asparagine was not detected strongly suggests that these asparagines are glycosylated. The C4-purified preparation was estimated to have an IG$ ₀ of about 5-10 nM in the hydrogen peroxide release assay (Example 1(E)).

Example 13

Determination of Specificity of the Neutrophil

Inhibitory Factor

To test the specificity of the Neutrophil Inhibitory Factor of the present invention, and to confirm that it did not inhibit neutrophil activation by a general cytotoxic mechanism, the activity of the inhibitor was assessed in a non-neutrophil cell adhesion-based assay, platelet aggregation. The effects of the hookworm Neutrophil Inhibitory Factor on blood platelet aggregation were examined. Platelet aggregation was performed with human platelet-rich plasma (PRP) . PRP was stirred at 37°C in an aggregometer (Scienco Model 247, Morrison, CO) and aggregation was initiated by the addition of _b0 μM ADP (Sigma, St. Louis, MO) . Aggregation was monitored as a change in light transmittance, and is expressed as the initial rate of aggregation. A concentration of Neutrophil Inhibitory Factor of approximately 150 nM, a concentration that completely blocked neutrophil function as assessed by neutrophil-HUVEC and neutrophil-plastic adhesion assays, homotypic neutrophil aggregation and hydrogen peroxide release by neutrophils, had no inhibitory effect on ADP-induced aggregation of human platelets.

Example 14

CDllb/CDl8 Integrin is a Primary Receptor for Neutrophil

Inhibitory Factor from Hookworm

(A) Immunopreciptation of ¹²⁵ I-Labelled NIF Using Monoclonal Antibodies to CDllb/CD18 in the Presence of Neutrophil Extract.

NIF purified from Ancylostoma canirium was radiolabeled using the following method. Approximately 30 μg NIF was labeled with 2 mCi Na ¹²⁵ I (carrier free; Amersham, Arlington Hills, IL) using Enzymobeads

(BioRad, Hercules, CA) Briefly, to a 1.5 ml eppendorf test tube was added 360 μl of the Enzymobead suspension together with 180 μl of a 1% beta-D-glucose solution, NIF and Na ¹²⁵ I. This mixture was allowed to react at room temperature for 30 minutes. Labeled NIF was separated from unbound ¹²⁵ I-iodine by desalting on a PD10-DG column (BioRad, Hercules, CA) using phosphate buffered saline (0.1 M sodium phosphate pH 7.2, 0.15 M sodium chloride) containing 1% bovine serum albumin as elution buffer. Radioactive fractions containing NIF were pooled. The specific activity of the ¹²⁵ I-NIF was 13.9 μCi/μg. Various leukocyte proteins were assessed for ability to capture NIF in immunoprecipitation experiments. Potential cellular receptors for NIF were selected from a detergent extract of leukocytes using specific monoclonal antibodies.

Leukocytes were prepared from human blood using Mono-poly (ICN, Biomedicals Inc., Costa Mesa, CA) . The leukocyte cell pellet was resuspended in 1 ml resuspension buffer (20 mM Tris pH 7.5, 150 mM NaCl, 1 mM CaCl ₂ ) followed by the addition of 1 ml extraction buffer (2% Triton X-100, 20 mM Tris pH 7.5, 150 mM NaCl, 1 mM CaCl ₂ ) . Cells were incubated on ice 30-60 minutes, vortexing briefly every 10 minutes. Cell debris was pelleted at 5000 g for 5 minutes at 4°C

Monoclonal antibody-test protein complexes were formed by incubating 10 μg specific monoclonal antibody with 200 μl of leukocyte detergent extract at 4°C for 4 hours. To this mixture was added 2.5 μl of the ¹²⁵ I-NIF and these reagents were incubated at 4°C for 18 hours.

Precipitation of the complex was effected by adding this mixture to a 1.5 ml eppendorf test tube containing 50 μl of protein G-sepharose (Pharmacia, Pistdcaway NJ; resuspended in TACTS 20 buffer (0.05% Tween 20, 20 mM Tris pH 8, 120 mM NaCl, 2 mM CaCl ₂ ) with 1% bovine serum albumin) and gently agitating at 4°C for 2 hours.

The protein G-sepharose beads were subsequently washed four times with TACTS 20 buffer. Fifty microliters of Laemmli sample buffer (Laemmli, U.K., 1970, Nature, 227:680-685) containing 5% /3-mercaptoethanol was then added to the aspirated beads; this material was incubated at 100°C for 10 minutes and loaded onto 4-12% gradient SDS-polyacrylamide gels (Novex, San Diego, CA) . Gels were dried after running and visualized by exposure to X-Omat film (Kodak, Rochester, NY) in the presence Quanta III screens

(Dupont, Wilmington, DE) at -70°C Size standards were ¹⁴ C-Rainbow markers (Amersham, Arlington Hills, IL) .

When monoclonal antibodies (MAb) directed to the CDllb/CD18 integrin complex (OKM-l, ATCC# CRL8026; LM-2, ATCC# HB204) were used in these experiments, ^I-NIF was precipitated as evidenced by a band that migrated with an apparent molecular weight of approximately 41,000 daltons upon autoradiography. Precipitation of ^12S I-NIF was dependent on the presence of these antibodies as well as the presence of leukocyte extract. Furthermore, the precipitation of ¹⁵ I-NIF was not observed in the presence of a one hundred fold molar excess of cold NIF. ¹²⁵ I-NIF did not precipitate when MAbs to other leukocyte integrins were used including MAbs directed against the VLA-4 (L25.3; Becton Dickinson, Sunnyvale, CA) and CDllc/CDlδ (SHCL-3; Becton Dickinson, Sunnyvale, CA) integrin complexes. A relatively minor amount of ¹²⁵ I-NIF was observed when a MAb directed against the CDlla/CD18 (TS1/22; ATCC# HB202) integrin complex was used. This was likely due to cross-reactivity of the anti-CDlla/CD18 antibody with the related integrin complex CDllb/CD18. These results demonstrate that CDllb/CD18 is a cell-surface receptor f r Ancylostoma caninum NIF on leukocytes.

(B) Precipitation of ¹²⁵ I-CDllb/CD18 Using Biotinylated NIF

As another approach to identify NIF receptors on leukocytes, biotin-labeled NIF was used to precipitate NIF-associating proteins from a detergent extract of surface iodinated leukocytes.

NIF was biotinylated by conjugation to its carbohydrate moieties. Approximately 16 μg of NIF purified from hookworm (Ancylostoma caninum) lysates (hydroxyapatite eluate; see Example 2(D)) was oxidized with 50 mM NaI0 ₄ in 1 ml 0.1 M sodium acetate, pH 5.5. After 20 minutes at 4°C the reaction was terminated with the addition of 100 μl 165 mM glycerol. Oxidized NIF was separated from other reaction products using a Microcon 10 concentrator (Amicon, Beverly, M* , and diluted into 100 μl 0.1 M sodium acetate, pH 5.5. Biotinylation was effected by the addition of 400 μl 6.25 mM biotin-LC-hydrazide (Pierce, Skokie, IL) . The reaction was allowed to proceed for 18 hours at 4°C Biotinylated NIF was worked up by buffer exchange into phosphate buffered saline (PBS; 0.1 M sodium phosphate, 0.15 M sodium chloride, pH 7.2), using a Microcon 10 concentrator. To 250 μl of the concentrate was added an equal volume of glycerol, giving a final NIF-biotin concentration of approximately 32 μg/ml. This material was stored at -20°C

The anti-CD18 integrin complex monoclonal antibodies LM-2 and OKM-l (anti-CDllb/CD18; ATCC #HB204 and CRL8026, respectively) and TS1/22 (anti-CDlla/CD18; ATCC# HB202) were biotinylated using the protocol described above.

Cell surface iodination of human leukocytes was done using the following procedure. A -total leukocyte fraction, prepared from 90 ml of fresh human blood using Mono-Poly density gradient separation (ICN Biomedical, Costa Mesa, CA) , was suspended in 0.5 ml phosphate

buffered saline. To the cell suspension was added 2 mCi Na ¹²⁵ I (carrier free; Amersham; Arlington Heights, IL) , 60 μl 0.03% hydrogen peroxide and 100 μl lactoperoxidase at 2 mg/ml (BioRad; Hercules, CA) . The reaction was allowed to proceed for 30 minutes at room temperature, with gentle agitation every two minutes. The reaction was terminated by the addition of 25 mM KI in PBS, and the cells were washed two times with PBS. The leukocyte cell pellet was resuspended in 1 ml resuspension buffer and leukocyte extract was prepared as described above in Example 14(A) .

Sixty microliters of NIF-biotin (32 μg/ml) was diluted with 40 μl resuspension buffer and incubated with 200 μl ¹²⁵ I-labeled leukocyte extract at room temperature for 6 hours. Precipitation of

NIF-associating proteins from the leukocyte extract was effected by the addition of 100 μl streptavidin-agarose (Pharmacia; Piscataway, NJ) to this mixture. Test tubes were agitated gently for 18 hours at 4°C Beads were subseguently washed four times with 500 μl TACTS-20 buffer (0.05% Tween 20, 20 mM Tris pH 8, 120 mM NaCl, 2 mM CaCl ₂ ) , and associated proteins were solubilized with 50 μl sample buffer (5% /3-mercaptoethanol) and analyzed by SDS-PAGE as described in Example 5. Control precipitations were performed in a similar manner with biotinylated monoclonal antibodies to CDllb/CD18 and CDlla/CDlδ.

Biotinylated NIF precipitated two ¹²⁵ I-labeled polypeptides that, when separated by 6% SDS-PAGE, had apparent molecular weights of about 170 kDa and about 95 kDa. These polypeptides comigrated on SDS-PAGE in this experiment with the two polypeptides that were precipitated by the anti-CDllb/CD18 monoclonal antibodies LM-2 and OKM-l. This data strongly suggests that CDllb/CD18 is a major receptor for NIF on leukocytes when considered with the results of the

previous experiment (Example 14(A)), in which CDllb/CDl8 was shown to associate with NIF.

Example 15 Preparation Of Native Neutrophil Inhibitory Factor From Toxocara canis

(A) Preparation of Toxocara Lysate.

Frozen canine worms Toxocara canis were obtained from Antibody Systems (Bedford, TX) and were stored at -70°C until homogenized. Toxocara canis were homogenized on ice in homogenization buffer [0.02 M Tris-HCl pH 7.4, 0.05 M NaCl, 0.001 M MgCl ₂ , 0.001 M CaCl ₂ , 1.0 X 1(T ⁵ M E-64 Protease Inhibitor (CAS 66701-25-5), 1.0 X 10 ^"6 M pepstatin A (isovaleryl-Val-Val-4-amino-3-hydroxy-6-methyJ--heptanoyl -Ala-4-amino-3-hydroxy-6-methylheptanoic acid, CAS 26305-03-3), 1.0 X 10" ⁵ M chymostatin (CAS 9076-44-2), 2.0 X 10" ⁵ M APMSF (amidinophenylmethylsulfonyl fluoride-HCl) , 5% (v/v) glycerol] using an Ultra-Tarrax homogenizer (Janke and Kunkel, Stanfen, Germany) . The protease inhibitors E64, pepstatin A, chymostatin, and APMSF were obtained from Calbiochem (La Jolla, CA) . Approximately 3-6 ml of homogenization buffer was used to homogenize each gram of frozen worm. Twenty-four grams of worms was used in total. Insoluble material was pelleted by two sequential centrifugation steps: 40,000 X g--^ at 4°C for 25 minutes followed by 105,000 X g-,^ at 4°C for 1 hour. The supernatant solution was clarified by passage through glass wool and a 0.45 μm cellulose acetate filter (CoStar, Cambridge, MA) .

(B) Concanavalin A Sepharose Chromatography of Toxocara Lysate

Toxocara canis lysate (66 ml) was absorbed to 26 ml of Concanavalin A Sepharose (Pharmacia, Piscataway, NJ) pre-eguilibrated with Con A buffer [0.02 M Tris-HCl, pH

7.4, 1 M NaCl, 0.001 M CaCl ₂ , 0.001 M MnS0 ₄ ] by recycling it through a 1.6 X 13 cm column at a flow rate of 4 ml/minute (119 cm/hour) for 2 hours. The column was at room temperature (24°C) while the reservoir of lysate was maintained on ice throughout the procedure. The column was subsequently washed with 100 ml of Con A buffer. Material that had activity in anti-adhesion assays (see, Section (D) below) was eluted with approximately 3-5 column volumes of Con A buffer containing 0.5 M methyl-alpha-mannopyranoside (CAS

617-04-09) at a flow rate of 1 ml/minute (30 cm/hour) . The eluted material was concentrated to 5 ml using an Amicon stirred ultrafiltration vessel equipped with a 10,000 molecular weight cutoff membrane, then diluted to 50 ml with deionized water, and reconcentrated to 2.3 ml using a centrifugal ultrafiltration unit with a 10,000 molecular weight cut-off (Polysciences, Inc., Warrington, PA) Material used for molecular sieve chromotography with Superdex columns (1.5 ml) was additionally concentrated to 0.5 ml using centrifugal ultrafiltration units with a 10,000 molecular weight cut-off (Amicon, Inc., Beverly, MA).

(C) Molecular Sieve Chromatography Using Superdex 200 HR. Material eluted from immobilized Concanavalin A

(see step (B) above) and concentrated by ultrafiltration was loaded on a 1.0 cm X 30 cm column of Superdex 200 HR

(Pharmacia, Piscataway, NJ) . The column was pre-equilibrated with 0.01 M potassium phosphate, pH 7.35, and 0.15 M NaCl at 24°C The chromatography was conducted at a flow rate of 0.25 ml/minute. Anti-adhesion activity eluted with an apparent molecular weight of approximately 20,000.

(D) Assay of Neutrophil Inhibitory Activity Isolated

From Toxocara canis

Material eluted from Concanavalin A Sepharose with methyl alpha-mannopyranoside was assayed by the neutrophil-HUVEC adhesion assay (see Example 1(B)) and was found to inhibit the adhesion of neutrophils to endothelial cells. Adhesion inhibitory activity was also demonstrated using the neutrophil-plastic adhesion assay. (Example 1(C)). Material purified by chromatography on both

Concanavalin A Sepharose and Superdex 200 HR inhibited neutrophil adhesion in the neutrophil-adhesion assay (see Example 1(C)).

Example 16 In Vivo Characterization Of Neutrophil Inhibitory Factor

Neutrophil Inhibitory Factor isolated from canine hookworms was tested in an animal model of acute inflammation.

Peritoneal inflammation was induced in 150-250 gram Sprague-Dawley rats by an intraperitoneal injection of nine ml of 2% oyster glycogen in H ₂ 0 (see Baron et al., Journal of Immunological Methods. 4jJ:305, 1982; McCarron et al., Methods in Enzvmology. 108:274, 1984; Feldman et al., Journal of Immunology. 113:329, 1974; Rodrick et al., Inflammation. .6:1, 1982; and Kikkawa et al., Laboratory Investigation. 3Q:76, 1974).

NIF was prepared as described in Example 2. Lysate from approximately 20,000 hookworms (48.2 g wet weight) was prepared and chromatographed on ConA, Superdex, and hydroxyapatite (HA) . The active fractions from two equivalent HA runs were combined to yield 41 ml of HA material. One ml of NIF solution (11 μg) was administered simultaneously with the glycogen by the intraperitoneal route or thirty minutes prior to glycogen administration by the intravenous route. Four

hours later the peritoneal exudate was harvested by purging the peritoneal cavity with 30 ml of Hanks Balanced Salt Solution without Ca ⁺⁺ or Mg ⁺⁺ , supplemented with 0.03% EDTA and blood cells were counted on a Celldyn 3000 (Abbott Laboratories, North Chicago, IL) automated multiparameter differential cell counting instrument. The major cellular component in the exudate was neutrophils. Figure 10 depicts the effects of varying doses of Neutrophil Inhibitory Factor isolated from canine hookworms on neutrophil infiltration in peritoneal inflammation in rats induced by interperitoneal infusion with glycogen. Glycogen (9 ml) and Neutrophil Inhibitory Factor (1 ml) were injected simultaneously by intraperitoneal route. Figure 10 shows the results of six independent experiments. NIF caused a dose dependent inhibition of neutrophil infiltration to the rat peritoneal cavity in response to glycogen.

A second study was performed to determine if intravenous administration of NIF could prevent glycogen-induced rat peritoneal inflammation. In one set of rats, NIF and glycogen were administered by the intraperitoneal route as previously described. In a second group of rats, 1 μg of NIF was administered intravenously thirty minutes prior to the intraperitoneal infusion of glycogen. A third group of animals received glycogen and NIF treatment was replaced with saline. Four hours later the peritoneal exudate was collected and blood cells were counted. Figure 11 depicts the effect of Neutrophil

Inhibitory Factor isolated from canine hookworms on neutrophil infiltration in peritoneal inflammation in rats induced by intraperitoneal infusion of glycogen. Neutrophil Inhibitory Factor (1 ml) was injected by intraperitoneal route in conjunction with intraperitoneal infusion of glycogen, or by intravenous

route thirty minutes prior to infusion of glycogen. Figure 11 represents a summary of the six independent experiments for the intraperitoneal administration of NIF and the results of the single experiment for the intravenous administration of NIF. These results demonstrate that NIF, when administered by either the intraperitoneal or intravenous route, was effective in the prevention of peritoneal inflammatory response in glycogen-stimulated rats.

Example 17

Inhibition of Neutrophil-Mediated Inflammation In Vivo bv Recombinant Neutrophil Inhibitory Factor

The in vivo anti-inflammatory properties of recombinant NIF (rNIF) were tested in a rat ear inflammation assay (adapted from Young et al., 1984).

In this assay, inflammation was induced in the rat ear by topical administration of arachidonic acid. Sprague-Dawley rats (250g) were anesthetized with pentobarbital (initial dose of 65 mg/kg intraperitoneal; Anpro Pharmaceutical, Arcadia, CA) ; rats were maintained at a surgical plane of anesthesia for the duration of the experiment (4 hours) . A catheter was inserted into the femoral vein of the anesthetized rat. One hundred microliters of recombinant NIF (produced in Pichia pastoris; see Example 12) at a concentration of 20 mg/ml in PBS was injected via the catheter. Control rats received 100 μL sterile 0.14 M NaCl. Five minutes after the IV administration of rNIF, arachidonic acid (Sigma, St. Louis, MS; diluted 1:1 with acetone to a final concentration of 500 mg/ml) was applied to the right ear in three 10 μl applications each to the ijnside and the outside of the ear. The right ear thus^received a total dose of 30 mg arachidonic acid. The left ear, used as a background control, received a total of 60 μl acetone.

Four hours after administration of arachidonic acid the rat was sacrificed with C0 ₂ .

Neutrophil infiltration into the arachidonic acid-treated ear tissue was quantitated indirectly by determining myeloperoxidase activity. A tissue sample was obtained from the center of each ear using a 7 mm skin punch (Miltex; Lake Success, NY) . The tissue sample was cut into small pieces and added to a 16 x 100 mm test tube that contained 0.5 ml HTAB buffer (0.5% hexadecyltrimethylammonium bromide in 50 mM sodium phosphate, pH 6.4; HTAB was purchased from Sigma, St. Louis, MO) . The ear tissue was homogenized for 20 seconds using an Ultra-Turrax (Janke and Kunkel; Staufen, Germany) at high speed. Insoluble matter was removed from the homogenate by centrifugation. at 14,000 x g for 10 minutes followed by filtration through Nytex gauze. Myeloperoxidase determinations were done in triplicate in 96 well polystyrene plates (Costar; Cambridge, MA) . Twenty five microliters of HTAB-solubilized ear tissue was added to each well, and to this was added 100 μl of substrate solution. Substrate solution comprised two components: (1) 0.012% H ₂ 0 ₂ in 0.1 M sodium acetate pH 4.5 and (2) 0.3 mg/ml 3,3' ,5,5'-tetramethylbenzidine in 10% HCl, combined immediately prior to use at a ratio of 0.125:1. After ten minutes the reaction was stopped by the addition of 125 μl 1 M H ₂ S0 ₄ . Samples were quantitated colorimetrically at 450 nm and background was read at 650 nm. A standard curve was generated using human leukocyte myeloperoxidase (Sigma; St. Louis, MO) .

Recombinant NIF had a protective effect on arachidonic acid-induced neutrophil infiltration into ear tissue. Figure 12 shows that ear t ssue from rats that received rNIF had a mean of 1.6 myeloperoxidase units/ml (MU/ml) whereas ears from rats that received saline had a mean of 4.1 MU/ml, when background

myeloperoxidase activity is subtracted (n=10 in each group) . One myeloperoxidase unit will produce an increase in absorbance at 470 nm of 1.0 per minute at pH 7.0 and 25°C, calculated from the initial rate of reaction using guaiacol as substrate (Desser, R.K. , et al., Arch. Biochem, Biophys. 14δ.:452 (1972)). Neutrophil infiltration was thus reduced -60% in rats that received rNIF (8 mg/kg IV) ; there is a significant difference at the 95% confidence level between rats that received NIF and rats that received saline (Student's t test) . These results are consistent with the demonstration that hookworm-derived NIF prevented neutrophil infiltration into the peritoneal cavity of rats in response to glycogen (see Example 16) . These data further provide evidence that rNIF acts.as a potent anti-inflammatory agent in vivo.

Example 18

The Use of Neutrophil Inhibitory Factor DNA Sequences to Isolate Neutrophil Inhibitory Factor-Related Proteins NIF cDNA sequences are used as probes to isolate

DNA sequences that encode proteins that are functionally and structurally related to NIF.

Genomic DNA or cDNA libraries are formed using standard procedure (for example see Molecular Cloning. A Laboratory Manual. Sambrook, J., Fritsch, EF., and

Maniatis, T. 2nd Ed. Cold Spring Harbor Laboratory Press, CSH, NY 1989) . These libraries may be from any animal, fungal, bacterial or viral source, such as Ancylostoma caninum. other Ancylostoma species, other helminths and mammals including human placental tissue.

Such libraries are screened for useful clones by nucleic acid hybridization using NIF cDNA sequences isolated from Ancylostoma as probe. For example, NIF cDNA fragments of about 100-2000 base pairs labeled for detection by standard procedure (for example, see

Molecular Cloning. A Laboratory Manual. Sambrook, J., Fritsch, EF., and Maniatis, T. 2nd Ed. Cold Spring Harbor Laboratory Press, CSH, NY 1989) is hybridized with a library from another tissue or another species under conditions of variable stringency. More preferably, however, reduced stringency hybridization conditions are utilized (eg 6X SSC [SSC is 150 mM NaCl, 15 mM trisodium citrate], 0.02 M sodium phosphate pH 6.5, 5X Denhardt's solution, 0.5% (w/v) SDS, 0.01 M EDTA, 100 μg/ml sheared, denatured salmon sperm DNA, 0.23% dextran sulfate, 20-30% formamide at 42°C for 18 hours) . Also, more preferably, reduced stringency conditions are used to wash filters after hybridization (0.5 to 2X SSC at 45-60°C for 20 minutes after two prewashes with 2X SSC for 15 minutes) .

NIF-related complementary DNAs isolated using the techniques described above are subjected to nucleotide sequence analysis using the procedure of dideoxy sequencing (Sanger et al, 1977, Proc. Natl. Acad. Sci. USA 24:5463-5467). Isolates containing open reading frames (i.e., initiating with a methionine and terminating with a TAA, TGA or TAG stop codon) are inserted into suitable vectors for protein expression in either bacterial, yeast, insect or mammalian cells. Expression systems comprise vectors designed to secrete recombinant protein (i.e., fusion of cDNA isolate open reading frame with a known secretion signal sequence for that cell type) into the culture medium. Vectors lacking a homologous secretion signal sequence are also used for expression. Either conditioned media or cell lysate, depending on the expression system used, is tested for inhibitory activity using one or more of the following criteria for neutrophil activaftion: release of hydrogen peroxide, release of superoxide anion, release of myeloperoxidase, release of elastase, homotypic neutrophil aggregation, adhesion to plastic surfaces,

adhesion to vascular endothelial cells, chemotaxis, transmigration across a monolayer of endothelial cells and phagocytosis.

Proteins that are structurally related to NIF and that are inhibitory in one or more of these neutrophil function assays would be considered to belong to the NIF family of related molecules.

Example 19

Expression of Functional Recombinant NIF in E. coli DNA for the NIF-IFL coding region, initiating at the codon that corresponds to the N-terminal methionine, is inserted into an ______ coli expression vector. Examples of such vectors are given in Balbas, P. and Bolivar, F. , 1990 (Methods in Enzymology, 185:14-37) . The DNA is inserted into the E. coli expression vector using methods similar to the methods of insertion of the NIF-IFL coding region into mammalian and yeast expression vectors described in Examples 11 and 12, respectively. PCR oligonucleotide primers are designed to generate an amplification product that contains the NIF-IFL coding region. As was described in connection for the methods for insertion of NIF-IFL into mammalian and yeast expression vectors (see Examples 11 and 12, respectively) , primers are engineered so that this fragment contains 5' and 3' restriction sites that are compatible with insertion into the selected expression vector. The expression construct is preferably engineered so that the recombinant NIF will be secreted into the cytoplasm and not the periplasmic space. This may be accomplished by omitting an ______ coli secretion signal from the construct.

______ coli cells are transformed with«*the NIF-IFL expression vector construct using standard methods. (See, e.g.. Molecular Cloning A Laboratory Manual, Sambrook, J. Fritsch, E.F. and Maniatis, T. , Second

Edition, Cold Spring Harbor Laboratory Press, 1989, 1.74-1.84). Cells are grown in appropriate media (e.g. Luria Broth; see Molecular Cloning. A Laboratory Manual, Sambrook, J. Fritsch, E.F. and Maniatis, T., Second Edition, Cold Spring Harbor Laboratory Press, 1989, A.l) and harvested before they reach the stationary phase of growth.

The majority of the recombinant NIF should be present in the cytoplasm in the form of insoluble and functionally inactive aggregates. The solubilization and refolding of the recombinant protein present in these aggregates may be accomplished using known methods such as those reviewed in detail in Kohno et al., 1990 (Methods in Enzymology, 185:187-195). Refolded recombinant NIF may be separated from unfolded recombinant NIF and other reaction products using a number of standard chromatographic techniques, including C4 reverse phase HPLC (see, e.g.. Example 2(E) ) . Refolded recombinant NIF is tested for functional activity using the neutrophil function assays described in Example 1.

This recombinant NIF is not glycosylated.

Example 20

Preparation of Functional Recombinant NIF by Refolding 'Insoluble' Methionyl-NIF Produced in the E. coli

Cytoplasm

(A) Description of the E. coli expression vector pMa5-NIl/3

PCR oligonucleotide primers were designed to generate an amplification product that contains the

NIF-IFL coding region. The PCR product initiates at the first Asn-codon of mature NIF-IFL which^as a result of the amplification was changed from AAT to AAC The 3'-primer replaces the TGA translational stop codon by a TAA triplet and introduces a Spel, a Hindlll and a Bglll

site downstream of the coding region. The PCR primers were as follows:

Pst414:

5'-CCTCCTCCTA-GATCTAAGCT-TACTAGTTTA-TAACTCTCGG-AATCGATAA A-ACTC (54-mer; 3'-primer matching with the C-terminus)

Pst415:

5'-AACGAACACA-ACCTGAGGTG-CCCG (24-mer; 5'-primer matching with the N-terminus)

After digestion with Hindlll, the correctly sized PCR fragment was isolated from agarose-gel and inserted on an E. coli expression vector downstream of the phage lambda P _R promoter. The recipient vector was.opened with Ncol, treated with DNA polymerase I (Klenow fragment) and subsequently digested with Hindlll. The resultant vector, designated pMa5-NIl/3, is schematically shown in Figure 14; the sequence of the relevant part of the vector is shown in Figure 15. The NIF-IFL region present in this vector was entirely sequenced to rule out the presence of unwanted mutations. The expression module consists of the following elements: (1) The phage lambda P _R promoter. (2) A small cistron which is present upstream of the Met-NIF-IFL region; this upstream cistron includes the first nine codons of the phage lambda cro gene and terminates at the TAA stop located in between the Shine Delgarno (SD)-box and the ATG initiator codon of Met-NIF-IFL (see Figure 14 and 15) . The leader-cistron has the potential to code for a 31 residue polypeptide. Such a two-cistron arrangement, in addition to being found in a number of 'natural' operons, has been used succesfully for .improving the expression of heterologous genes whose level of expression is thought to be limited by the initiation of translation (Schoner et al., PNAS 1:5403-5407 (1984);

Spanjaard et al., Gene |SC):345-351 (1989); Makoff and Smallwood, NAR lδ:1711-1716 (1990). (3) An open reading frame encoding Methionyl-NIF-IFL. The construction scheme is indeed such that the NIF-FL1 5'-end is correctly fused to an ATG initiator codon. Upstream of this ATG codon a SD-sequence (eg, GGAGGT; see Figures 14 and 15) is present. (4) Two copies of a phage fd derived transcription terminator (fdT) downstream of the Met-NIF-IFL coding region.

(B) Production of 'insoluble' methionyl-NIF.

To assess the effectiveness of pMa5-NIl/3 in expressing the NIF-IFL gene, the vector was introduced in W3110 cells harboring pcI857. The plasmid pcI857 specifies resistance to kanamycin (20 μg/ml) ,. encodes a temperature sensitive repressor of the lambda P _R promoter, and is compatible with the pMa5-NIl/3. Cultures were grown in LB medium at 2δ"C to a density of about 2 x 10 ⁸ cells/ml and then induced at 42 ^* C for 2-3 hours. Analysis of total cellular extracts of induced and non-induced cells by SDS-PAGE indicated that a new -33 kDa protein (calculated molecular weight of NIF-IFL = -29 kDa) is synthesized upon thermo-induction of the promoter. In addition to total cellular extracts, we also analysed the pellet (insoluble) and supernatant (soluble) fraction obtained by opening the induced cells by sonication and clearing the lysate by centrifugation. The results indicated that the newly synthesized -33 kDa protein precipitates intracellularly, i.e. forms so-called inclusion bodies. Following fractionation on an SDS-polyacrylamide gel, transfer onto ProBlott (ABI) and visualization by coomassie-staining, the -33 kDa band was excised and its N-terminal amirtό acid sequence determined. The sequence obtained was: M-N-E-H... . This result demonstrated that the initiator methionine was not removed from the primary translation product and

clearly identified the 33 kDa band as recombinant NIF. It was estimated that the recombinant NIF protein accumulates to -10 mg per liter and per OD _6J0 unit.

(C) Renaturation of NIF protein expressed in E. coli . W3110 E___, coli cells containing the plasmid pMa5-NIl/3 were grown in a shake flask incubator in six liter flasks each containing 1.5 liters of LB media at 28°C until the optical density (OD) was in the range of 0.6-0.9 au at 550 nm. An additional 1.5 liters of LB media at 56°C was added to each flask to induce expression of recombinant NIF, and the flasks were incubated at 42'C with shaking. The OD was monitored and cells were harvested by centrifugation when the OD was within the range of 1.0-1.5. The cell pellets were frozen at -80 ^β C Each tube contained about 3.5 g cells. Fifteen milliliters of TES buffer (0.05 M Tris, 0.05 M sodium ethylenediaminetetraacetate, 15% (w/v) sucrose, pH 8.0) was added to one tube, and the tube was sonicated to thaw and disperse the cell pellet. The suspension was then distributed into two 30 ml glass centrifuge tubes. An additional 2.5 ml of TES was used to wash the original tube and this wash solution was added to the glass tube. The suspensions in the glass tubes were sonicated (Branson Sonic Power Co., Danbury, Connecticut) four times for 30 seconds each, with an ice incubation between sonications to maintain the temperature <10°C throughout the procedure. The tubes were then centrifuged at 10,000 rpm for 20 minutes (12,100 x g,, _^ ) at 4°C The supernatants were discarded. The pellets were resuspended in 15 ml of PSX buffer

(0.02 M potassium phosphate, 1 M sodium chloride, 1% (v/v) Triton X-100, pH 7.2) per tube and ^' sonicated at a low setting to break up the pellets followed by a 15 second sonication at medium power. The tubes were cooled on ice, resonicated at medium power for 15

seconds, and then centrifuged at 10,000 rpm for 20 minutes (12,100 x g,^) . The supernatant was discarded. The entire PSX resuspension/centrifugation process was repeated two additional times. The pellets were resuspended in 15 ml of PBS buffer (0.01 M sodium phosphate, 0.15 M sodium chloride, pH 7.3) per tube, briefly sonicated, and centrifuged as before. The PBS resuspension was repeated one additional time. The pellets were then resuspended in PBS by sonication, 12.5 ml per tube. The contents of both tubes were combined, and the volume was brought to 30 ml by the addition of further PBS. The protein concentration of the purified inclusion bodies was determined using the DC Protein Assay (Bio-Rad, Hercules, California) .

An aliquot of purified inclusion bodies containing 5 mg of protein was placed in each of several 1.7 ml plastic microcentrifuge tubes. The tubes were microcentrifuged for 10 minutes at 4 ^* C The supernatants were discarded. Each pellet was resuspended in 1 ml of 0.05 M Tris, 1 % (w/v) Sarkosyl, pH 7.5 using sonication.

The tubes were vortexed at 37 ^* C for 48 hours using a Thermomixer vortexer (Eppendorf, Hamburg, Germany) . Following this incubation, samples were submitted for NIF activity assays (see Example 1) . Typically, activity corresponding to the activation (refolding) of 3% of the NIF present was found.

Example 21

Isolation and Characterization of NIF from Ancylostoma caninum

(A) Cloning and Seguencing of NIF sequences from A. caninum

Two new full-length coding regions that code for proteins related to NIF-IFL (see Example 10) were identified by PCR technology using single stranded oligonucleotide DNA primers that match with the NIF-IFL N- and C-terminal ends. These primers were as follows:

YG1:

5'-ATG-GAG-GCC-TAT-CTT-GTG-GTC-TTA (5'-primer matching with the N-terminal region encoding: M-E-A-Y-L-V-V-L)

YG2: 5'-TCA-TAA-CTC-TCG-GAA-TCG-ATA-AAA-CTC (3'-primer matching with C-terminal sequence corresponding to: E-F-Y-R-F-R-E-L-Stop codon)

The YG1/YG2 primer couple was found to yield a correctly sized PCR product when using a total RNA preparation of Aj_ caninum (see Example 10) as template. First strand cDNA synthesis (First-Strand cDNA Synthesis Kit of Pharmacia, Uppsala, Sweden; 10 pmoles of the YG2 primer; -15 μg of total RNA) and the subsequent amplification by PCR were carried out according to the manufacturer's specifications. The PCR was carried out with Taq DNA polymerase (Boehringer, Mannheim, Germany) , 100 pmoles of both YG1 and YG2 and using 30 temperature cycles (1 minute denaturation step at 95 ^* C; 1 minute annealing period at 55'C; 1.5 minute elongation step) . The obtained PCR product was isolated from agarose gel and cloned onto a phagemid vector (allowing preparation of single stranded DNA) . Three clones, designated PCR-NIF5, PCR-NIF7 and PCR-NIF20, were retained for sequence determination. PCR-NIF5 was found to be identical to NIF-IFL. PCR-NIF7 and PCR^IF20, however, represent two new NIF sequences. Their sequences are shown in Figure 16.

Additional full-length A^. caninum NIF sequences were isolated by screening a cDNA library using as hybridization probe a radiolabeled PCR fragment obtained with primers that target sequences which are well conserved among the seven Ai. caninum NIF sequences described in Example 10 (IFL, 3P, 2FL, 3FL, 4FL, 6FL and IP) . The following primers were used:

YG3:

5'-CAC-AAT-GGT-TAC-AGA-TCG-AGA-CTT-GCG-CTA-GGT-CAC the 5'-primer targeting the region which in NIF-IFL encodes the amino acid sequence H-N-G-Y-R-S-K-L-A-L-G-H)

YG4:

5 ^/ -T-TTT-TGG-GTA-GTG-GCA-GAC-TAC-ATG the 3'-primer targeting the region which in NIF-IFL encodes H-V-V-C-H-Y-P-K-(I)

Poly(A+) RNA was prepared from adult worms using the QuickPrep mRNA Purification Kit (Pharmacia, Uppsala, Sweden) . Using this poly(A+) RNA preparation as template, an amplification product of about the expected length was obtained with the YG3/YG4 primer couple (the PCR conditions were as descibed above) . The amplification product was shown by gel-electrophoretic analysis to be rather heterogeneous with respect to length. The YG3/YG4 primers were indeed designed to target sequences that flank that part of the coding region where the various NIF sequences display significant differences in length; the heterogeneous nature of the PCR product indicated that the primers are useful for the amplification of several different isoforms. The PCR product were gel-purified and subseguently radiolabeled by "random primer extension" ( ^" "QuickPrime Kit™; Pharmacia, Uppsala, Sweden) for use as hybridization probe. A cDNA library was constructed

using described procedures (Promega Protocols and Applications Guide 2nd Ed.; Promega Corp.). About 3 μg of mRNA was reverse transcribed using an oligo(dT)-Notl primer-adaptor [5'-TCGCGGCCGC(T) _1S ; Promega Corp., Madison, WI] and AMV (Avian Myeloblastosis Virus) reverse transcriptase (Boehringer, Mannheim, Germany) . The enzymes used for double stranded cDNA synthesis were the following: E. coli DNA polymerase I and RNaseH from BRL Life Technologies (Gaithersburg, MD) and T4 DNA polymerase from Pharmacia. The obtained cDNA was treated with EcoRl methylase (RiboClone EcoRl Linker Ligation System; Promega) . The cDNAs were digested with Notl and EcoRl, size selected on a 1% agarose gel (fragments of between 1000-7000 base-pairs were eluted using the Geneclean protocol, BIO101 Inc., La. Jolla, CA) , and unidirectionally ligated into the EcoRI-NotI arms of the lambda gtll Sfi-Not vector (Promega) . After in vitro packaging (Gigapackll-Gold, Stratagene, La Jolla, CA) recombinant phage were obtained by infecting strain Y1090 (Promega) . The usefulness of the cDNA library was demonstrated by PCR analysis (Taq polymerase from Boehringer; 30 temperature cycles: 1 minute 95°C; l minute 50"C; 3 minutes 72 ^* C) of a number of randomly picked clones using the lambda gtll primer #1218 (New England Biolabs, Beverly, MA) in combination with the above mentioned oligo(dT)-Notl primer adaptor. The majority of the clones was found to contain cDNA inserts of variable size.

Approximately lxlO ⁶ cDNA clones (duplicate plaque-lift filters were prepared using Hybond™-N;

Amersham, Buckinghamshire, England) were screened with the radiolabeled YG3/YG4 PCR fragment using the following prehybridization and hybridization conditions: 5X SSC (SSC: 150 mM NaCl, 15 mM trisodium citrate) , 5X Denhardt's solution, 0.5% SDS, 50% formamide, 100 μg/ml sonicated fish sperm DNA (Boehringer) , overnight at

42"C. The filters were washed 4 times in 2X SSC, 0.1% SDS at 37°C After overnight exposure to X-ray film, numerous plaques that hybridized to the probe were identified; it was estimated that about 0.1-0.2% of the clones scored positive. After a second hybridization round (24 positives were analyzed at lower plaque-density so as to isolate single pure clones) , a number of phage clones were subjected to PCR anlysis and those cDNA inserts which were found to be large enough to encompass the entire coding region were subcloned as Sfil-NotI fragments on pGEM-type phagemids (Promega) . We have determined the sequence of eight of these NIF cDNAs (i.e., AcaNIF3, AcaNIF4, AcaNIF6, AcaNIF7, AcaNIF9, AcaNIFlδ, AcaNIF19, and AcaNIF24) . The data are shown in Figure 16.

(B) Expression of Functional NIF Proteins from A. caninum in Pichia Pastoris.

The segments of DNA encoding AcaNIF24, AcaNIF6, AcaNIF4, and AcaNIF9 were PCR amplified using pGEM-type vectors containing the respective cDNAs (see above) as template. The 5'-primers contained no restriction sites and matched with the 5'-end of that part of the coding regions corresponding to the mature protein. Also, the first codon was altered from AAT to AAC (both codons translate to asparagine) . The sequences of the 5'-primers for the various NIF sequences were as follows:

AcaNIF24: 5'-AAC-GAA-CAC-AAC-CTG-ACG-TGC-CC AcaNIFδ: 5'-AAC-GAA-CAC-AAA-CCG-ATG-TGC-CAG-C AcaNIF4: 5'-AAC-GAA-CAC-AAA-CCG-ATG-TGC-GAG

AcaNIF9: 5'-AAC-GAA-CAC-GAC-CCA-ACG-TGΦ-CC

The 3'-primers were composed of δ codons at the 3'end of the coding region, a TAA stop replacing the TGA

stop of the natural gene, and three unique restriction endonuclease sites (Spel, Hindlll, and Bglll) . The 3'-primers used were:

AcaNIF24: 5'-CCT-CCT-CCT-AGA-TCT-AAG-CTT-ACT-AGT-TTA-AAA-TCG-ATA-A

AA-CTC-CTT-GCT-ATC

AcaNIF6:

5'-CCT-CCT-CCT-AGA-TCT-AAG-CTT-ACT-AGT-TTA-TAA-CTC-TCG-G

AA-TCG-ATA-AAA-CTC AcaNIF4:

5'-CCT-CCT-CCT-AGA-TCT-AAG-CTT-ACT-AGT-TTA-TAA-CTC-TCG-G

AA-TCG-ATA-AAA-CTC

AcaNIF9:

5'-CCT-CCT-CCT-AGA-TCT-AAG-CTT-ACT-AGT-TTA-TAG-CTC-TCG-A AA-CGG-ATA-AAA-ATA

Amplification was accomplished using 100 pmoles of each primer, 2 units of Vent polymerase in lx Vent buffer (New England Biolabs, Beverly, MA), 0.2 mM of each dNTP and 100 ng of template DNA. The PCR conditions were the same for all twenty cycles: denaturation at 95'C for 1 minute, primer annealing at 60"C for 1 minute, and amplification for 1.5 minutes at 72°C The amplification product was gel-purified and digested with Spel. The amplification product was ligated into Stul-Spel cleaved pHIL7SP8 using standard methods. The ligation mixture was used to transform E. coli WK6 selecting for ampicillin resistant clones. In each case a correct clone was identified by restriction and DNA sequence analysis. These plasmids, designated pYAM7SP-AcaNIF24, pYAM7SP-AcaNIF6, pYAM7SP-AcaNIF4, and pYAM7SP-AcaNIF9, were used to transform-^the P. pastoris yeast strain GTS115(his4) , as described in Example 12 (B) . Selection of His ⁺ transformants and subsequent selection for NIF expression were performed as described

in Example 12(B). The accumulation of functional NIF protein in Pichia cell supernatant was detected and quantified using the LM2/CDllb/CDlδ based ELISA with 3D2-HRP detection (Example 1) in the case of AcaNIF24, AcaNIF6, and AcaNIF4 and using the competitive assay for LM2/CDllb/CD18 (Example 1) in the case of the AcaNIF9.

(C) Purification and Characterization of NIF proteins AcaNIF24. AcaNIF6. AcaNIF4. and AcaNIF9. The functionally active recombinant AcaNIF24, AcaNIFδ, and AcaNIF4 proteins were purified and characterized in greater detail. Following methanol induction for 48 hours, Pichia cell supernatants were obtained by centrifugation for 15 minutes at 1,800 x g. In the case of the recombinant proteins AcaNIF24 and AcaNIF6, the ,250 ml supernant was adjusted to pH 7.0 by adding Tris-HCl and kept overnight at 4'C Precipitated material was removed by centrifugation. The cleared supernatant was loaded on a 3D2-immunoaffinity resin and bound material eluted with glycine-HCl pH 2.5 (Example 27). The eluted fractions were neutralized and concentrated by ultrafiltration. Both proteins were found to migrate as a single band on SDS-PAGE (4-20% gradient gel; Novex) . Edman degradation confirmed that correctly processed proteins were produced. The following N-terminal amino acid sequences were found:

AcaNIF24: N-E-H-X-L-T-X-P-Q-N

AcaNIF6: N-E-H-K-P-M-X-Q-Q-X-E-T-E-M-P where X represents an unidentified residue.

The concentrations were determined^spectrophoto- metrically and the samples were assayed in the plastic adhesion and peroxide release assays (see Example 1) . Both recombinant AcaNIF24 and AcaNIF6 proteins were

found to be equally active as recombinant NIF-IFL (Figure 17) .

Recombinant AcaNIF4 was purified by hydroxyapatite and reverse-phase chromatography essentially as described in Example 23, however, the gel filtration step on Superdex was omitted. The purified protein was found to migrate as a single band on SDS-PAGE (4-20% gradient gel; Novex) . The concentration was determined spectrophotometrically. The results obtained in a competitive binding assay with biotinylated NIF-IFL indicated that AcaNIF4 had a significantly lower affinity for the LM2/CDllb/CD18 complex than NIF-IFL (Figure 18) .

Recombinant AcaNIF9 was partially purified by reverse-phase chromatography (see Example 23) - Edman degradation revealed N-E-H-D-P as N-terminal amino acid sequence confirming that correctly processed AcaNIF9 protein was produced. This partially purified protein was found to have a considerably higher mobility on SDS-PAGE (4-20% gradient gel; Novex) than the

Pichia-produced NIF-IFL protein (30-35 kDa compared to 40-80 kDa) consistent with the presence of seven N-glycosylation sites in NIF-IFL and of only two potential N-glycosylation sites in AcaNIF9. The sample containing AcaNIF9 was tested in the competitive binding assay described in Example 1. The results demonstrated that binding of biotinylated recombinant NIF-IFL to the LM2/CDllb/CD18 complex can be prevented by recombinant AcaNIF9.

Example 22

Isolation and Characterization of a NIF Protein from Ancylostoma ceylanicum.

(A) Cloning and Seguencing of NIF Sequences From A. ceylanicum.

A full-length A. ceylanicum NIF gene was isolated by screening a cDNA library using as hybridization probe a PCR fragment effected from the same species. The PCR fragment was obtained using primers that target sequences which are highly conserved among the seven A. caninum NIF isoforms described in Example 10 (IFL, 3P, 2FL, 3FL, 4FL, 6FL and IP) . These primers, designated YG3 and YG4, are described in Example 20.

Poly(A-t-) RNA was prepared from A. ceylanicum adult worms using the QuickPrep mRNA Purification Kit

(Pharmacia, Uppsala, Sweden) . Using this poly(A--) RNA preparation as template, an amplification product of about the expected length (i.e., about the same length as the PCR fragment seen with A. caninum RNA as template) was obtained with the YG3/YG4 primer, couple. First strand cDNA synthesis (First-Strand cDNA Synthesis Kit; Pharmacia) and the subsequent amplification by PCR were carried out according to the manufacturer's specifications using 10 pmoles of the YG2 primer (see Example 20) and 100 ng of A. ceylanicum mRNA. The PCR was carried out with Taq DNA polymerase (Boehringer, Mannheim, Germany) , 100 pmoles of both YG3 and YG4 and using 30 temperature cycles (1 minute denaturation step at 95°C; 1 minute annealing period at 55°C; 1.5 minutes elongation step) . The PCR product was gel-purified and subsequently radiolabeled by "random primer extension" ("QuickPrime Kit™; Pharmacia) for use as hybridization probe.

An A. ceylanicum cDNA library was constructed in lambda gtll using the procedures described in Example 20. The quality of the cDNA library was demonstrated by PCR analysis (Taq polymerase from Boehringer; 30 temperature cycles: 1 minute at 95 ^* C; 14ιinute at 50"C; 3 minutes at 72°C) of a number of randomly picked clones using the lambda gtll primer #1218 (New England Biolabs) in combination with an oligo(dT)-Notl primer adaptor

(Promega) . The majority of the clones were found to contain cDNA inserts of variable size.

About 5xl0 ⁵ lambda cDNA clones were screened with the radiolabeled YG3/YG4 PCR fragment using the hybridization conditions described in Example 20.

Approximately 60 positives were identified. The cDNA insert of one positive clone, shown by PCR analysis (see above) to contain an insert of sufficient size to encompass the entire NIF coding region (-850 bp) , was transferred to pGEM-9Zf(-) (Promega) as a Sfil-NotI fragment and its sequence determined. The sequence of the cDNA clone, designated AceNIF3, is shown in Figure 19.

(B) Expression of a NIF like Protein from A.-ceylanicum in a Phaqe-Attached Form.

The A. ceylanicum AceNIF3 region coding for the mature protein was cloned onto a phage display vector according to the procedures described for NIF-IFL in Example 22. The N-terminal amino acid sequence of the authentic A. ceylanicum NIF protein is not known; it is, therefore, difficult to unambiguously locate the secretion signal processing site on the deduced amino acid sequence (Figure 19) . The following oligonucleotide primers were chosen to PCR amplify the

AceNIF3 coding region:

YG16:

5'-GTCGCAACTG-CGGCCCAGCC-GGCCATGGCC-GCTGACGAAC-CAACGTGCA

A-GCAG (54-mer; 5'-primer; the Ncol site and AceNIF3 N-terminus are underlined); and

YG15:

5'-GAGTTCTCGA-CTTGCGGCCG-CACCTCCGAT-AGGTGGATAA-CGGAGTGA

(48-mer; 3'-primer; the Notl site and AceNIF3 C-terminus are underlined) .

Following Ncol/Notl digestion, the PCR product was gel-purified and cloned between the Ncol and Notl sites of the recipient vector. The resultant vector, designated pAN-AceNIF3, contains the intended in-frame fusion of the pelB, AceNIF3, and M13 gill coding regions.

Phages displaying the AceNIF3 protein were obtained by infecting TGl(su ⁺ ) bacteria harboring pAN-AceNIF3 with M13-VCS 'helper'-phage (see procedures described in Example 22) . The rescued phages, resuspended in PBS, were assayed by an ELISA on immobilized LM2/CDllb/CD18 complex (see Figure 20) . Phages displaying the A _^ . caninum NIF-IFL isoform (Example 22) were used as positive control. Following a 30 minute incubation with varying amounts of Pichia produced recombinant NIF-IFL, 10 ¹⁰ virions displaying either recombinant NIF-IFL or recombinant AceNIF3 were added to the LM2/CDllb/CD18 coated wells. After a 90 minute incubation period, the amount of bound phages was detected with rabbit anti-phage serum and goat anti-rabbit alkaline phosphatase conjugate. Binding of phages to the immobilized receptor (see Figure 20) clearly indicates that the AceNIF3 protein must be displayed in a functionally active form on the phage surface. The data given in Example 22, show that phage binding occurs only when they display the NIF protein, i.e. non-displaying control phages do not bind to the LM2 monoclonal antibody nor do they bind to the LM2/CDllb/CD18 complex.

Displacement of both NIF-IFL- and AceNIF3-displaying phages by an increasing amount of soluble Pichia produced recombinant NIF-IFL demonstrates that both NIF proteins bind to the same site on the CDllb/CD18 receptor with a comparable affinity. Phage display of the AceNIF3 protein was also demonstrated by Western

blot. After fractionation on an SDS-10% polyacrylamide gel, phage proteins were transferred onto ProBlott membrane (Applied Biosystems Inc.) and incubated consecutively with a rabbit anti-pglll serum (GATC GmbH, Konstanz, Germany) and goat anti-rabbit alkaline phosphatase conjugate. Bands corresponding to the wild type phage pglll protein and to the NIF-pglll fusion product could be visualized.

(C) Construction of pYAM7SP-AceNIF3 and Expression in Pichia.

The segment of DNA encoding AceNIF3 was PCR amplified from a subclone of AceNIF3 in pGEM-9Zf(-) (see above) using unique primers for the 5'- and 3'-ends of the coding region. The 5'-end of the proteolytically processed AceNIF3 being not unambiguously defined, a hybrid 5'-end was created based on sequence homology between AcaNIF9 and AceNIF3: the three N-terminal codons of proteolytically maturated AcaNIF9 were used as 5'-end, followed by six codons originating from the AceNIF3 sequence. The resulting N-terminal amino acid hybrid sequence was: N-E-H-E-P-T-C-K-Q, while the natural AcaNIF9 sequence was N-E-H-D-P-T-C-P-Q, and the natural AceNIF3 sequence was K-G-D-E-P-T-C-K-Q. The sequence of the 5'-primer used was 5'-AAC-GAA-CAC-GAA- CCA-ACG-TGC-AAG CAG. The 3'-primer was composed of 8 codons at the 3'-end of the coding region, a TAA stop replacing the TGA stop of the natural gene, and three unique restriction endonuclease sites (Spel. Hindlll, and Xbal) . The sequence of the 3'-primer used was 5'-CCT-CCT-CCT-TCT-AGA-AGC-TTA-CTA-GTT-TAG-ATA-GGT-GGA-T

AA-CGG-AGT-GAC-G.

Amplification was accomplished using 100 pmoles of each primer, 2 units of Vent polymerase in lx Vent buffer (New England Biolabs, Beverly, MA), and 0.2 mM of each of dATP, dCTP, dGTP, and dTTP. One hundred

nanograms of pGEM-9Zf(-) containing AceNIF3 were used as template DNA. The PCR conditions were the same for all twenty cycles: denaturation at 95"C for 1 minute, primer annealing at 60°C for 1 minute, and amplification for 1.5 minutes at 72'C. The amplification product was gel-purified and digested with Spel.

The amplification product was ligated into Stul-Spel cleaved pHIL7SP8 using standard methods. The ligation mixture was used to transform E. coli WK6 selecting for ampicillin resistant clones. Based on restriction and DNA sequence analysis, a correct insert sequence in one of the resulting plasmid clones, pYAM7SP-AceNI3, was selected to transform the P. pastoris yeast strain GTS115(his4) , as described in Example 12(B) . Selection of His ⁺ transformants and subsequent selection for AceNIF3 expression were performed as described in Example 12(B) . The presence of AceNIF3 in Pichia cell supernatant was detected and quantified in a competitive binding assay with biotinylated NIF1 (Example 1) .

Following methanol induction for 4δ hours, Pichia cell supernatant was obtained by centrifugation. The crude supernatant was shown to inhibit the adhesion of human neutrophils to plastic.

Example 23

Production by E. coli of Functionally Active NIF as Either a Bacteriophage-Attached Form or as 'Free' Soluble Protein (A) Cloning of NIF-IFL on a phage display vector. A phagemid-vector was assembled in which the

NIF-IFL region coding for the mature protein is fused at its N-terminus to the secretion signal sequence derived from the pelB gene and at its C-terminal end to the filamentous phage M13 gene III (gill) . This gene fusion was placed under the transcriptional control of the Plac

promoter. Some of the pelB codons were replaced by synonymous triplets so that the secretion signal contains an Ncol restriction site. An extra Ala-codon was introduced between the pelB and NIF-IFL regions such that the junction matches more closely the prokaryotic prototype signal sequence processing site. The NIF-IFL and pglll (product of gill) encoding regions are separated by (i) a linker sequence in which a Notl site is embedded and (ii) a TAG (amber) triplet which serves as a translational stop codon in a su ^' strain but is frequently read as a sense codon in su ⁺ bacterial cells.

A schematic representation of the phagemid vector, designated pAN-NIF-lFL, is shown in Figure 21. pAN-NIF-lFL was constructed by (i) PCR-amplification of the NIF-IFL coding region with primers that contain 5'-extensions whose sequence is such that (ii) the Ncol/Notl directional cloning of the gel-purified PCR fragment in the recipient vector results in the intended in frame fusion of the pelB, NIF-IFL, and gill coding elements.

The NIF-IFL coding region was PCR-amplified making use of the following two oligonucleotide primers:

LJ045:

5'-GTCGCAACTG-CGGCCCAGCC-GGCCATGGCC-GCTAATGAAC-ACAACCTGA G-GTGC (54-mer; 5'-primer; The Ncol site and NIF-IFL N-terminus are underlined) LJ046:

5'-GAGTTCTCGA-CTTGCGGCCG-CAGGTGGTAA-CTCTCGGAAT-CGATAAAAC T-C (51-mer; 3'-primer; the Notl site and NIF-IFL C-terminus are underlined)

The NIF-IFL region and flanking sequences present in pAN-NIF-lFL were entirely sequenced to rule out the presence of unwanted mutations.

(B) Display of functional NIF bv filamentous phages.

In su ⁺ bacteria such as TGI, the pAN-NIF-lFL phagemid-vector has the potential to code for a NIF-lFL-pglll fusion protein. When the TGl[pAN-NIF-lFL] host cells are infected with a so-called 'helper'-phage, this fusion protein can, during morphogenesis, become incorporated into filamentous virions (both 'helper'-phages and pseudo-virions which encapsidate one specific strand of the phagemid) . Phage particles were rescued with M13-VCS 'helper'-phage (Stratagene) infection as follows. A 1 ml culture of TGl[pAN-NIF-lFL] grown at 37'C in 2xTY (2xTY: Tryptone 16 g/L; Yeast extract 10 g/L; NaCl 5 g/L) containing 100 μg/ml carbenicillin (or ampicillin) and 1% glucose to a density of ODgoo _nπ - - 0.5-0.6 is infected with M1-3-VCS at a multiplicity of infection of -20. The infected culture is incubated at 37°C for 30 minutes without shaking and then for another 30 minutes with shaking. A 10 ml prewarmed 2xTY aliquot containing both carbenicillin (100 μg/ml) and kanamycin (50 μg/ml) is inoculated with the 1 ml infected culture. The mixture is incubated with shaking first for 60 minutes at 37°C and then overnight at -30°C After removal of the infected cells by centrifugation 1:5 volume 20% polyethyleen glycol/2.5 M NaCl is added to the supernatant. Following a 60 minute incubation on ice, the precipitated phages are collected by centrifugation and resuspended in PBS (Na ₂ HP0 ₄ .2H ₂ 0 1.14 g/L; KH ₂ P0 ₄ 0.2 g/L; NaCl 8.0g/L; KC1 0.2 g/L; pH 7.3) . The rescued phages were shown to display functionally active NIF in several assays: (A) Western blot (After fractionation by SDS-10% PAGE, phage proteins were transferred onto ProBlott (Applied Biosystems Inc. , Foster City, CA) membrane and incubated with rabbit anti-phage serum and goat anti-rabbit alkaline phosphatase conjugate. A band corresponding to

the NIF-pglll product could be visualized) ; (B) CDllb/CDlδ-ELISA (see Figure 22) (NIF-phage were then assayed for binding to CDllb/CDlδ (non-displaying phage were used as negative control) . CDllb/CD18-coated wells were prepared either by direct immobilization using 0.25 μg/ml immunopurified CDllb/CDlδ receptor (Diamond et al., 1990, J. Cell Biol., ill, 3129-3139), or by immuno-capture with monoclonal antibody LM2 (ATCC number: HB 204) . Binding of phages was detected with rabbit anti-phage antiserum and goat anti-rabbit alkaline phosphatase conjugate. Binding of phages to the immobilized receptor was shown to occur only when they display the NIF protein. It was also shown that NIF-phage are not able to bind to the LM2 monoclonal antibody nor to the CDllb/CD18-coated wells after a pre-incubation with ImM Pichia-produced recombinant NIF-IFL for 30 minutes) ; (C) 3D2-ELISA (3D2 is a non-neutralizing mouse monoclonal antibody specific for NIF (see Example 26) . In contrast to non-displaying control phages, NIF-phage were found to bind to

3D2-coated wells. NIF-phage binding could be eliminated by either blocking the 3D2-wells with 1 mM Pichia-produced recombinant NIF-IFL or blocking the NIF-phages with 1 mM 3D2 monoclonal antibody) ; and (D) Panning against CDllb/CDlδ (pAN-NIF-lFL phage (10 ¹⁰ virions) were mixed with an equal amount of irrelevant non-displaying phage (fd-tet; 10 ¹⁰ virions) , diluted in 100 μl Binding Buffer (PBS, 1 mM CaCl ₂ , 1 mM MgCl ₂ , 0.4% Tween-20 and 2% Skim-Milk) and incubated in a CDllb/CD18-coated microtiter-well. After incubation for

120 minutes, and washing with PBS containing 1 mM CaCl ₂ 1 mM MgCl ₂ and 0.4% Tween-20 ten times, bound phage were eluted during a 10 minute incubation with glycine-HCl pH 2.0. Following neutralization with 1 M Tris-HCL pH 8.0, the number of tetracycline resistant (fd-tet) and ampicillin resistant (pAN-NIF-lFL) colony forming units

was determined; pAN-NIF phage were 30-fold enriched over the non-displaying fd-tet phage) .

The above experiments, e.g. functional display of NIF-IFL on filamentous phage and the specific enrichment of such NIF-phages by binding selection, show it is possible to use the phage technology for the identification of higher-affinity NIF variants (i.e., both naturally occuring isoforms or engineered mutants) . Similar to what has been done in the immunoglobulin field, it should be possible to clone the vast majority of the A. caninum NIF protein repertoire on phage and then to select the highest affinity NIF isoform by subjecting the phage library to several consecutive binding selection cycles (panning) using the CDllb/CD18 receptor as target. Our sequence data show that the extent of conservation of the 5' and 3' termini of the region encoding mature NIF allows the design of (degenerate) oligonucleotide primers to rescue a substantial part of the NIF protein repertoire. For example, we generated PCR-amplification fragments of the expected length using a lambda DNA preparation of the pooled A. caninum cDNA library as target with the following oligonucleotide primer sets: 5'-primer targeting the N-terminus: an equimolar mixture of three primers that contain at their 3'-end the following matching sequences: AAT-GAA-CAC-AAC-CTG-ASG-TGC-3' AAT-GAA-CAC-GAC-CCA-ACG-TGT-3' AAT-GAA-CAC-AAA-CCG-ATR-TGC-3' where S = C or G and R = A or G; and

3'-primer targeting the C-terminus: an equimolar mixture of two primers that contain at their 3'-end the following matching sequences: TAA-CTC-TCG-GAA-TCG-ATA-AAA-3' TAA-CTC-TCG-AAA-CSG-ATA-AAA-3' where S = C or G.

The PCR primers contain 5'-extensions which incorporate restriction sites allowing the facile unidirectional cloning of the amplification product in an appropriate display vector.

(C) Secretion of Soluble and Functionally Active NIF. The phagemid display vector containing the NIF gene, pAN-NIF-lFL, is suitable for the production of NIF-IFL in both a phage-attached form and as 'free' soluble protein. In su ^" bacteria such as WK6, the pAN-NIF-lFL phagemid-vector has the potential to direct the synthesis of the NIF-IFL protein in a 'free' (i.e., not phage-attached) form.

Overnight induction of the Plac promoter by addition of isopropyl-/S-D-thiogalactopyranoside (l mM final concentration) to a WK6[pAN-NIF-lFL] culture was found to result in the accumulation of CDllb/CD18-binding activity as shown by ELISA (on LM2/CDllb/CDlδ plates; detection was done with HRP-conjugated monoclonal antibody 3D2) . The recombinant NIF-IFL protein could be detected in both the supernatant of the induced culture and in a total cell lysate prepared by sonication of the induced cells followed by a clearing step. Comparison of the ELISA-signal with that generated by known amounts of Pichia-produced rNIFl allowed us to estimate that the rNIF protein accumulates to about lmg per liter of E. coli culture. The rNIF protein was also immunopurified on a 3D2-Emphaze column (see Example 27) starting from a French-Press lysate of induced WK6[pAN-NIF-lFL] cells. Material eluted at low pH was still active as determined by the LM2/CDllb/CDlδ ELISA. The E. coli rNIF protein was shown to migrate as a sharp band on SDS-polyacrylamide gel and could be detected by rabbit anti-Pichia-rNIFl serum in immuno-blot analysis.

Example 24

Alternative Purification Method For Pichia Produced NIF.

Cell-free supernatant was filtered (0.2 μm) and submitted to a diafiltration on a polyethersulfone omega membrane (30kDa cut-off; 0.75 ft ² ; Filtron) with 10 volumes of 50 mM citric acid pH 3.5 containing ImM EDTA. After adjustment to pH 7.4 by adding 1 M Tris-HCl, the solution was left on ice for at least one hour. Precipitated material was removed by filtration (0.2 μm) . The cleared supernatant was submitted to a second dialfiltration (10 volumes 10 mM phosphate pH 7.4). Afterwards calcium chloride was added to a final concentration of 0.3 mM. The solution was applied on a MacroPrep (40 μm) Hydroxyapatite (Bio-Rad Laboratories) column equilibrated with 10 mM phosphate pH 1,4 and containing 0.3 mM CaCl ₂ . After washing with 5 column volumes of the equilibration buffer, the recombinant NIF protein was eluted with 90 mM phosphate pH 7.4. Fractions containing recombinant NIF were identified by binding assays on LM2/CDllb/CDlδ plates and pooled. Subsequently, the protein present in the pooled fractions was further purified by reversed phase chromatography on a Poros Rl/H (Perseptive Biosystems) column equilibrated with 10 mM ammonium formate pH 6.4 and 10% acetonitrile. Recombinant NIF was eluted by increasing the acetonitrile concentration. The fractions containing NIF were identified by gel-electrophoresis and were pooled. Acetonitrile present in this pool was removed in a rotavapor before freeze-drying. The dry protein was redissolved in PBS and applied on a Superdex

200 (Pharmacia) gel filtration column equilibrated in PBS. Fractions containing the NIF protein were pooled and concentrated by ultrafiltration on <an omega membrane (10 kDa cut-off; Filtron) . The recombinant NIF protein was stored at -80°C

Example 25

Expression of Functional Derivatives of NIF-IFL in Pichia pastoris.

(A) pMa5-hNIFl and pMc5-hNIFl Expression Constructs. The segment of DNA encoding NIF was PCR amplified from a subclone of NIF-IFL in Bluescriptll (Stratagene, La Jolla, CA) using unique primers for the 5'- and 3'-ends of the coding region.

The 5'-primer was composed of two restriction sites (EcoRl and Hpal) and the 23 first nucleotides of the region beginning at the 5'-end of proteolytically processed NIF and the succeeding δ codons. The codon for the first residue of the mature NIF was altered from AAT to AAC (both codons translate to asparagine) and constitutes part of the Hpal restriction sites

(GTT/AAC) . The sequence of the 5'-primer used was

5'-CCG-GAA-TTC-GTT-AAC-GAA-CAC-AAC-CTG-AGG-TGC-CC. The

3'-primer has been described in Example 12(B).

Amplification was accomplished using 100 pmol of each primer, 2 units of Vent polymerase in IX Vent buffer (New England Biolabs, Beverly, MA), and 0.2 mM of each of dATP, dCTP, dGTP, and dTTP. One hundred nanograms of Bluescriptll-containing NIF-IFL were used as template DNA. The PCR conditions were the same for all twenty cycles: denaturation at 95°C for 1 minute, primer annealing at 60'C for l minute, and amplification for 1.5 minutes at 72"C. The amplification product was gel-purified and digested with EcoRl and Hindlll.

The amplification product was then ligated into EcoRl-Hindlll cleaved pMa5-δ and pMc5-8 respectively

[Stanssens et al., Nucl. Acids Res. 17: 4441-4454 (1989)], using standard methods. The ligation mixtures were used to transform competent E. cold' WK6 (Zell et al., (1987) EMBO J. , 6: 1809-1615). Cells resistant to ampicillin and chloramphenicol, respectively, were selected and obtained on appropriate plates. Based on

restriction and DNA sequence analysis, a correct insert sequence in each of the resulting plasmid clones, pMa5-hNIFl and pMc5-hNIFl, were selected.

(B) Construction of pMa5-hNIFl/ΔGll-5. The NIF-IFL protein contains seven potential N (Asparagine) -glycosylation sites (consensus N-X-T/S amino acid sequence) . pMa5-hNIFl/ΔGll-5 is a derivative of pMa5-hNIFl (see above) in which five potential N-glycosylation sites of NIF-IFL have been modified by substituting glutamine residues for each of the asparagine residues in the corresponding consensus sequences. These residues are Asn ¹⁰ , Asn ¹⁸ , Asn ⁸⁷ , Asn ¹¹⁰ , and Asn ¹³⁰ , where the number in superscript corresponds to the . mino acid residue number of NIF-IFL (see Fig. 8) .

Stepwise site-directed mutagenesis was performed following the methodology described in Stanssens et al., (1989) , Nucl. Acids Res. 12 4441-4454, and using the following oligonucleotides: (I) ΔG11:dCCGGGCATTTCGGTACCTTGCTGCGGGCACCTC,

(II) ΔG12:dCCTAATCGAGTCTTGGAACCCGGGCATTTCTGTTCC,

(III)ΔG13 :dAACTGTCCGAGCATTGTCGTGCACTCATGTAGGCGCTTTTTTC,

(IV) ΔG14:dCAGAGCTTCAGAGATCTGGTTTGAGTTTTCG, and

(V) ΔG15:dCTCCTTCTTTTGTTTTCTGCAGGTTGAAAGCCTC. In the first mutagenesis round, the oligonucleotides I, IV and V were annealed together to the single strand DNA (ssDNA) template pMa5-hNIFl to modify the corresponding glycosylation sites Gil, G14 and G15. A resulting plasmid clone having the three intended sites altered was then used to prepare ssDNA template for the next mutagenesis round in order to modify the G12 and G13 remaining sites tfsing the appropriate oligonucleotides (II and III) .

(C) Construction of pMa5-NIF-lFL/Δh.G16-7.

The strategy outlined in (B) above was performed in parallel to construct another NIF-IFL derivative, pMa5-NIF-lFL/ΔhG16-7, in which the potential N-glycosylation sites G16 and G17 of NIF-IFL have been modified by substituting glutamine residues for each of the asparagine residues in the corresponding consensus sequences. These residues are Asn ¹⁹⁷ , and Asn ²²³ . In addition, the Hpal restriction site (GTTAAC) present in the NIF-IFL coding sequence was removed by introducing a silent mutation at the appropriate position: the AAC codon for Asn ¹⁶⁶ was replaced by a AAT codon.

Stepwise site-directed mutagenesis was performed following the methodology described in Stanssens et al., (1989), Nucl. Acids Res. 12: 4441-4454, and using the following oligonucleotides:

(VI) ΔG16:dCGGCTGTCCTTCAGTTTTCTGTATTTTCGGGTAGTGGC,

(VII) ΔG17:dGGATCCGCAGACGTCGTTTGGTCTGCTTTTTTTG, and

(VIII) Δh :dCTCCCAAAGGGCAATTAACAACTGCGC

In the first mutagenesis round, the oligo- nucleotides VII and VIII were annealed together to the ssDNA template to modify the glycosylation site G17 and to remove the Hpal restriction site. A resulting plasmid clone harbouring the two intended sites altered was then used to prepare ssDNA template for the next mutagenesis round in order to modify the G16 remaining site using the appropriate oligonucleotide (VI) .

(D) Construction of pMa5-hNIFl/Δh.Gll-7.

The NIF-IFL derivative hNIFl/Δh,Gll-7 was constructed using standard methods by combining appropriate fragments prepared from the vectors pMa5-hNIFl/ΔGll-5 (prepared as in (B) above) and pMa5-NIF-lFL/Δh,G16-7 (prepared as in (c above) . The 361 bp Agel-Hindlll fragment prepared from the vector pMa5-NIF-lFL/Δh,G16-7, and containing the three substitutions described in (C) above, was cloned into

the large Aqel-Hindlll vector fragment prepared from the vector pMa5-hNIFl/ΔGll-5, replacing the corresponding 361 bp Agel-Hindlll wild type NIF-IFL fragment of this vector. The presence of the seven Asn/Gln substitutions as well as of the modified Hpal restriction site (Asn ¹⁶⁶ codon modification) in the resulting plasmid, pMa5-hNIFl/Δh,Gll-7, was confirmed by sequencing analysis of the complete NIF insert. A one base pair deletion in the NIF sequence (a missing G nucleotide in the Gly ²⁰¹ GGA codon) revealed by this sequence analysis was corrected by site directed mutagenesis using the oligonucleotide dGTAAATCGGCTGTCCTTCAGTTTTCTG.

(E) Construction of pYAM7SP-hNIFl/ΔGll-5 and- Expression in Pichia pastoris.

The segment of DNA encoding hNIFl/ΔGll-5 was PCR-amplified from a subclone of pMa5-hNIFl/ΔGll-5 following the methodology described in Example 12(B) and using the same set of primers. After purification, the amplification product was digested with Spel and ligated into Stul-Spel cleaved pHIL7SP8 using standard methods. The ligation mixture was used to transform ______ coli WK6, and ampicillin resistant clones were obtained on ampicillin plates. Based on restriction and DNA sequence analysis, a correct insert sequence in one of the resulting plasmid clones, pYAM7SP-hNIFl/ΔGll-5, was selected to transform the P. pastoris yeast strain GTS115(his4) , as described in Example 12(B). Selection of His+ transformants and subsequent selection for NIF-1FL/ΔG11-5 expression were performed as described in

Example 12 (B) . The presence of NIF-1FL/ΔG11-5 in Pichia cell supernatant was detected and quantified using the LM2/CDllb/CD18 based ELISA with 3D2-HRP detection (see Example 1) .

(F) Construction of pYAM7SP-hNIFl/Δh.Gll-7 and expression in Pichia pastoris. The Hpal-Spel fragment of DNA encoding hNIFl/Δh,Gll-7 was prepared from the vector pMa5-hNIFl/Δh,Gll-7 and ligated into Stul-Spel cleaved pHIL7SP8 using standard methods. The ligation mixture was used to transform I . coli WK6, and ampicillin resistant clones were obtained on ampicillin plates. Based on restriction and DNA sequence analysis, a correct insert sequence in one of the resulting plasmid clones, pYAM7SP-hNIFl/Δh,Gll-7, was selected to transform the P. pastoris yeast strain GTS115(his4) , as described in Example 12(B) . Selection of His+ transformants and subsequent selection for NIF-lFL/Δh,Gll-7 expression were performed as-described in Example 12(B). The presence of NIF-lFL/Δh,Gll-7 in Pichia cell supernatant was detected and quantified using the LM2/CDllb/CDlδ based ELISA with 3D2-HRP detection (see Example 1) .

(G) Purification and Characterization of Recombinant NIF-1FL/ΔG11-5 and Recombinant NIF-lFL/Δh.Gll-7. Following methanol induction for 4δ hours, Pichia cell supernatants were obtained by centrifugation for 15 minutes at 1,800 x g. Recombinant NIF-1FL/ΔG11-5 was purified exactly as described in Example 23. The recombinant NIF-IFL mutant was found to migrate with an apparent molecular weight of 36-50kDa on SDS-PAGE (4-20% gradient gel; Novex) under non-reducing conditions. The band is more discrete and has a significantly higher mobility than wild type NIF-IFL produced in Pichia. Recombinant NIF-lFL/Δh,Gll-7 was purified by hydroxyapatite and reverse-phase chromatography (see Example 23; the gelfiltration step on Superdex was omitted) . Under non-reducing conditions, the purified protein was found

to migrate as a single band on SDS-PAGE (4-20% gradient gel; Novex) , with an apparent molecular weight of about 30kDa. The observed higher mobility and apparent lesser heterogeneity of NIF-1FL/ΔG11-5 and NIF-lFL/Δh,Gll-7 compared to NIF-IFL is likely due to the relatively decreased extent of glycosylation of these mutants compared to the wild-type proein.

Both mutants were evaluated in the plastic adhesion assay (see Figure 23) . The results indicate that elimination of part or all of the seven potential

N-glycosylation sites does not affect the potency of the NIF-IFL molecule as measured in this in vitro assay.

Example 26

Preparation of Monoclonal Antibodies to NIF Hybridomas producing MAbs that bind to NIF were prepared from mice immunized with Pichia-produced recombinant NIF-IFL by previously described methods (H.R. Soule, E. Linder, T.S. Edgington, Proc. Natl. Acad. Sci. USA 80:1332 (1983); G. Kohler and C Milstein, Nature (London) 256:495 (1975)).

Female balb-c mice between the ages of 8 to 15 weeks were inoculated subcutaneously with 10 μg of recombinant NIF (from Example 12) in Complete Freund's Adjuvant, then 3 weeks later were inoculated subcutane- ously with 10 μg of the recombinant NIF in Incomplete

Freund's Adjuvant, then 2 weeks later were inoculated mterperitoneally with 10 μg of the recombinant NIF in 4 mg of alum, then three to four weeks later and also 4 days prior fusion were mterperitoneally innoculated with 10 μg of the recombinant NIF in saline. Mice producing polyclonal antibody to recombinant NIF were determined by immunoassay of their seruict ^' using ¹²⁵ I labeled recombinant NIF and immobilized goat anti-mouse IgG. Mice having a titer between 1:25,000 to 1:50,000 were selected. Spleen cells were isolated from the mice

and were fused with tumor cells in a 5:1 ratio of spleen cells to tumor cells. Hybridoma cells expressing monoclonal antibody binding to NIF were selected by the same immunoassy. Monoclonal antibody was expressed from the selected hybridoma cells by inoculation mterperi¬ toneally of 5 x 10° hybridoma cells into pristane-primed female balb-c mice.

One monoclonal antibody designated 3D2 was shown to bind both hookworm-derived NIF and Pichia-produced recombinant NIF-IFL by antigen capture assay (E. Harlow and D.P. Lane, Antibodies : A Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1988) , pp. 192-193).

Example 27 Coupling of 3D2 Murine Monoclonal Antibody to 3M Emphaze Biosupport Medium (A) Single Step Method.

Twenty eight milligrams of 3D2 antibody was concentrated to 0.5 ml and 4.5 ml of 0.6 M sodium carbonate, 0.1 M sodium citrate, pH 9.0 was added. Three hundred fifteen milligrams of dry Emphaze Biosupport Medium (3M, St. Paul, Minnesota) was added to the antibody solution, and mixed end-over-end for one hour. The Emphaze slurry was collected over a plastic frit, draining the remaining antibody solution. The resin was washed with 20 ml of lOmM sodium phosphate, 0.15 M sodium chloride, pH 7.3. The collected resin was then mixed end-over-end with 3.0 M ethanolamine, pH 9.0 for 2.5 hours. The resin was again collected on a frit, and washed with 10 mM sodium phosphate, 0.15 M sodium chloride, pH 7.3 until the solution flowing through the frit was pH 7.3. The resin was stored 4 ^' n lOmM sodium phosphate, 0.15 M sodium chloride, 0.05% sodium azide, pH 7.3, until use. Two millilters of 3D2/Emphaze resin

coupled using this procedure was found to bind 1.4 mg of purified NIF protein.

(B) Two Step Method.

Thirty milligrams of 3D2 antibody was dialyzed into 0.5 M Tris-HCl, pH 4.0, concentrated to a volume of 5.0 ml (6 mg/ml) , and chilled to 4 ^β C Three hundred fifteen milligrams of Emphaze Biosupport Medium was added to the antibody solution and mixed end-over-end for 10 minutes at 4'C Sodium sulfate was added to a concentration of 0.8 M (563 mg) and the Emphaze slurry was mixed end-over-end for an additional 10 minutes at 4'C. The pH was raised to 9.0 by dropwise addition of IM NaOH. The coupling reaction was allowed to proceed for 60 minutes at 4°C with end-over-end mixing. The. Emphaze slurry was collected over a plastic frit, and remaining antibody solution was drained. The resin was washed with 20 ml of lOmM sodium phosphate, 0.15 M sodium chloride, pH 7.3. The reaction was quenched by the addition of 6 ml of 1.0 M ethanolamine, pH 9.3, and the resin in ethanolamine was mixed end-over-end for 2.5 hours at 4°C The resin was again collected over a frit, and washed with 0.2 M sodium phosphate, 0.5 M sodium chloride, pH 7.3 until the solution flowing through the frit was pH 7.3. The resin was suspended in 0.01 M sodium phosphate, 0.15 M sodium chloride, 0.05% (w/v) sodium azide, pH 7.3, until use. A 2ml 3D2/Emphaze column coupled using this method typically bound 1.5 mg of purified NIF protein.

(C) Scale-Up of Coupling Reactions. Twenty milliliters of resin has been made using both coupling methods by simply scaling^'up all volumes and amounts 10 fold. Both methods yielded resin capable of binding NIF protein. A 20 ml portion of 3D2/Emphaze resin coupled by the Two Step Method bound 10 fold more

NIF protein than did 2 ml columns (15-19 mg versus 1.5 mg) , whereas 20 ml of 3D2/Emphaze resin coupled using the Single Step Method did not scale linearly (typically 5-8 mg NIF protein bound) . Thus, the Two Step Coupling Method was used for larger couplings.

A 150 ml Two Step coupling was performed in which 2.25 g of 3D2 antibody in 450 ml of 0.5 M Tris-HCl, pH 4.0 at 4°C was mixed with 19 g of Emphaze Biosupport Medium in a 1000 ml microcarrier spinner flask (Bellco Glass Inc., Vineland, New Jersey) for 10 minutes at 4°C. Forty two grams of sodium sulfate was added to the Emphaze slurry and this was stirred for an additional 10 minutes at 4'C. The pH was raised to 9.0 by dropwise addition of 1 M NaOH through the arms of the spinner flask. The reaction was allowed to proceed for 60 minutes at 4"C with stirring. The Emphaze slurry was collected over a 90 mm, 0.45 micron filter (Corning Glass Works, Corning, New York) , and remaining antibody solution was drained. The resin was washed with 1 liter of 0.01 M sodium phosphate, 0.15 M sodium chloride, pH

7.3. The reaction was quenched by the addition of 500 ml of 1.0 M ethanolamine, pH 9.3 to the resin. The quenching reaction was allowed to continue 2.5 hours at 4°C with stirring. The resin was again collected over a 90mm, 0.45 micron cellulose acetate filter, and washed with 0.2 M sodium phosphate, 0.5 M sodium chloride, pH 7.3 until the solution flowing through the filter was pH 7.3. The resin was suspended in 0.01 M sodium phosphate, 0.15 M sodium chloride, 0.05% sodium azide, pH 7.3, until use. A 2 ml portion of resin from this coupling bound 1.5 mg of NIF protein, the same amount of NIF protein bound by resin coupled in a 2 ml reaction.

Example 28

Purification of Recombinant NIF Protein Using 3D2/Emphaze Immunoaffinitv Chromatography Column

Cell supernatant containing recombinant NIF protein was filtered through a Sartobran PH 0.07 micron dead-end filter (Sartorius North America, Bohemia, New York) , and then concentrated 10-50 fold by tangential flow filtration using a Mini Crossflow System containing 10 kDa Minisart polysulfone membrane modules (Sartorius) . The concentrate was then diafiltered against five volumes of 0.01 M sodium phosphate, 0.15 M sodium chloride, pH 7.3 in the Mini Crossflow apparatus. Immediately before application to the 3D2/Emphaze column, the concentrate was filtered through a 90mm, 0.22 micron cellulose acetate filter (Corning). Approximately 150 mg of NIF protein was applied to a 400 ml 3D2/Emphaze column at 20 ml/min. The concentrate was washed from the column with 400 ml of 0.1 M sodium phosphate, 0.15 M sodium chloride, pH 7.3 at 20 ml/min. The column flowthrough was collected and retained. The column was then washed with 400 ml of 1 M NaCl and the wash was discarded. The recombinant NIF protein bound to the column was eluted by applying 800 ml of 0.1 M glycine, pH 2.5. After elution, the purified NIF protein from the column was brought to neutral pH by the dropwise addition of 1 M Tris base. The column was then re-equilibrated to loading conditions by passing 800 ml of 0.1 M sodium phosphate, 0.15 M sodium chloride, pH

7.3, through it until the pH of solution exiting the column was pH 7.3.

When approximately 1 g of NIF protein had been purified by the 3D2/Emphaze column, the protein was pooled and then concentrated using an Easyflow 20 kDA polysulfone concentration apparatus (Sartorius) . The concentrated protein was then applied at a flow rate of 10 ml/min. to a 60 cm x 600 cm Superdex* 200 prep gel filtration column (Pharmacia, Piscataway, New Jersey) equilibrated in 0.01 M sodium phosphate, 0.15 M sodium

chloride, pH 7.3. The only peak observed during elution (870 - 1050 ml) corresponds to NIF protein.

Example 29

Neutrophil Inhibitory Factor is an Inhibitor of Eosinophil Adhesion to Vascular Endothelial Cells

NIF was assayed for effect on adhesion of human eosinophils to cytokine-stimulated endothelial cells. Eosinophils were isolated from normal individuals as described by Moser et al (1992a) [J. Immunol. 149:1432-1438] . Isolated eosinophils were cultured in the presence of 10 pM GM-CSF and 10 pM IL-3 for 24 hours following the procedure of Moser et al (1992a) . Endothelial cells were harvested from umbilical cord veins, seeded in tissue culture flasks and transferred to 24-well plates as described by Moser et al (1992a) . The adhesion assay was carried out following the procedure described by Moser et al (1992a) . Briefly, human umbilical vein endothelial cell (HUVEC) monolayers were washed with Hank's balanced salt solution (HBSS) and preincubated with 500 μl of TNFα at a final concentration of 10 ng/ml for 4 hours at 37'C. Immediately before use in adhesion assays, HUVEC monolayers were washed. Next, 2.5 X 10 ^s eosinophils in 500 μl of HBSS containing 5 mg/ml of purified human albumin were layered onto the washed HUVEC monolayers.

After incubation for 30 minutes at 37°C and saturated humidity/5% C02, the 24-well plate was submerged three times in a bath of 300 ml PBS to remove loosely adherent eosinophils. Plates were dried at 4°C and the number of adherent neutrophils was quantitated by measuring peroxidase activity, as described in Moser et al, 1992b [Blood 79:2937] .

Recombinant NIF (rNIF) inhibited adhesion of GM-CSF/IL-3 primed human neutrophils to TNF-activated HUVEC monolayers, to a maximum of approximately 63%

inhibition at 100 nM rNIF. About 50% inhibition of adhesion was obtained in the presence of approximately 10 nM rNIF (see Figure 13).

Example 30 Binding of NIF to Leukocytes

The interaction of NIF with leukocytes was assessed by flow cytometry using biotinylated recombinant NIF.

Biotinylated rNIF was prepared as described above for derivatization of NIF purified from hookworm homogenates with the exception that the rNIF was derivatized at a final biotin-LC-hydrazide concentration of 0.14 mM. A buffy coat preparation of human leukocytes, suspended in 2% newborn calf serum, 1% NaN ₃ (PBS-NCS) , was incubated with biotinylated rNIF (0.6 μg/ml ) for 5 minutes at room temperature. Red blood cells were lysed by addition of 3 ml 150 mM NH„C1, 1% NaN ₃ for 5 minutes at room temperature. The cells were washed twice in 1 ml PBS-NCS and resuspended in 50 μl wash buffer (5% newborn calf serum, 0.1% NaN ₃ in PBS) containing 0.25 μg/ml streptavidin-phycoerythrin (Pharmingen) . After 15 minutes at 4 ^* C the cells were washed and resuspended in 0.5 ml wash buffer. Flow cytometry was performed with a FACScan ^* instrument (Becton Dickinson) using Lysys II ^* software (Becton Dickinson) . Leukocyte populations were electronically gated using cytograms of forward versus right angle light scatter. Background binding was determined using biotinylated BSA (Pierce) .

The binding of biotinylated rNIF to lymphocytes (panels A, D and G) , monocytes (panels B, E and H) and granulocytes (panels C, F and I) was analyzed by flow cytometry. (See Figure 24) . Cell types were electroni- cally gated by using cytograms of forward versus right angle light scatter. Panels A-C illustrate negative control reactions using biotinylated BSA, panels D-F show reactivity with biotinylated rNIF and panels G-I show reactivity with biotinylated rNIF in the presence of a 50-fold molar excess of underivatized rNIF. All reactions were developed with streptavidin- phycoerythrin. Cell number is given on the ordinate and fluorescence instensity on the abscissa.

Greater than 95% of the gated granulocytes and monocytes were found to bind biotinylated rNIF relative to the BSA-biotin control (Fig. 24) . In contrast, rNIF associated with a discrete minor population of peripheral blood lymphocytes. Cellular binding of NIF was specific because coincubation of biotinylated rNIF and a 50-fold molar excess of non-derivatized rNIF abolished the reaction with all three leukocyte cell populations (Fig. 24) . These data demonstrate that in each of these leukocyte populations there exist cells that specifically bind NIF.

Example 31

Direct concordance of NIF binding and CDllb/CD18 expression

Dual reaction flow cytometry experiments were done to investigate CDllb/CDlδ expression of the lymphocyte population that bound NIF, using a monoclonal antibody to CDllb/CDlδ (LM2; ATCC# HB204) . Leukocyte staining with biotinylated rNIF was done as described in Example 30. In these experiments LM2 (1 μg/ml) was incubated with the whole leukocyte preparation in addition to biotinylated rNIF. After red blood cell lysis and washing, leukocytes were reacted with FITC-conjugated goat (Fab') ₂ anti-mouse IgG (Caltag) at 35 μg/ml in addition to streptavidin-phycoerythrin.

Direct concordance of NIF binding and CDllb/CDlδ expression was deomnstrated by dual fluorescence analysis by flow cytometry of peripheral blood lymphocyte populations (electronically gated by using cytograms of forward versus right angle light scatter) for rNIF binding versus binding of anti-CDllb/CDlδ monoclonal antibody LM2. All cells reactive with rNIF were positive for CDllb/CDlδ. (See Figure 25) . These experiments demonstrated a direct concordance between lymphocytes that bound NIF and those that expressed CDllb/CDlδ (Figure 25) . These results provide strong evidence that the integrin CDllb/CDlδ is a binding site for NIF on leukocytes.

Example 32

NIF Binds the I-domain Portion of the CDllb Polypeptide

The I-domain (=A domain) portion of the CDllb/CD18 integrin comprises a segment of the CDllb polypeptide which is approximately bounded by the residues Cys ¹²⁸ and Glycine ³²¹ (Michishita, M. , Videm, V., and Arnaout, M.A. (1993) Cell 72,657-867). To demonstrate direct interaction between NIF and the I-domain peptide, we expressed the I-domain peptide as a recombinant soluble fusion peptide in E. coli cells. The fusion is a 8 amino acid tag peptide termed FLAG ^* - The PCR primers used to clone I-domain were based on amino acid sequences of the CDllb polypeptide (Gly ^lll -Ala ³¹⁸ ) . The primer sequences were 5'-CCA-AAG-CTT-GGA-TCC-AAC-CTA- CGG-CAG-CAG-CC-3' (primer 1) , and 5'-CCA-TCT-AGA-CGC- AAA-GAT-CTT-CTC-CCG-AAG-CT-3' (primer 2). Primer 1 contains an additional 5'-Hindlll restricion endonuclease site, and primer 2 contains an additional 3 ' -XbaI site. The primers were used pairwise with random-primed single-stranded cDNA (cDNA Synthesis Plus, Amersham) synthesized from human monocyte total RNA (see Example 10 for protocols) . The starting material was -0.5 g human monocytes. PCR conditions were: denaturation at 95 'C for 30 seconds, annealing at 45 °C for 30 seconds, elongation at 72 'C for 2 minutes, for 30 cycles. All PCR reagents were from Perkin-Elmer. PCR amplification produced a fragment containing a -615 base pair I-domain coding region with appropriate 5'- and 3'-restriction sites. The Hindlll/Xbal-digested

fragment was ligated into the Hindlll/Xjal-cleaved plasmid vector pFLAG-1 ^™ (International Biotechnologies, Inc) . The construct was used to transform Epicurean Coli SURE ^* competent cells (Stratagene) , which were plated on LB media agar plates containing 100 μg/ml ampicillin and 0.4% glucose for selection. Three milliter overnight cultures inoculated with transformed colonies were grown at 37^ in LB + 100 μg/ml ampicillin. Isolation of plasmid DNA from the overnight cultures was performed with a Magic Miniprep kit

(Promega) , and isolated plasmid DNA was restricted with Hindlll and Xbal . Digestion products of -615 and -5370 bp were observed, indicating the presence of the CDllb I-domain-encoding insert in the expression vector. This expression plasmid was termed PM1. Five hundred milliters of LB + 100 μg/ml ampicillin was inoculated with 5 ml of an overnight culture of E. coli transformed with the plasmid PM1. The 500 ml culture was grown at 37 °C to an ODj- _o of 0.400 (-120 minutes), induced with 1.7 mM isopropyl b-D-thiogalactopyranoside (Sigma), and grown for two additional hours at 37 °C Cells were harvested by centrifugation at 5,000 g for 5 minutes, 25 °C. The cells were then resuspended in 50 ml of an extraction buffer-1 (50 mM Tris, pH 8.0, 5 mM EDTA, 0.25 mg/ml lysozyme, and 50 μg/ml NaN ₃ ) , and allowed to incubate for 5 minutes at 25 °C . To this preparation was added 5 ml of extraction buffer-2 (1.5 M NaCl, 100

mM CaCl ₂ , 100 mM MgCl ₂ , 0.02 mg/ml DNase 1, and 50 μg/ml ovomucoid protease inhibitor) , and this suspension was allowed to incubate for 5 minutes at 25 "C. Cells were further lysed by sonication for 4 cycles of 30 seconds at 70 W on a Branson Sonic power sonifier. Insoluble cellular debris was separated by centrifugation at 25,000 g for 60 minutes at 4 ^* C The cell lysate containing CDllb I-domain fusion peptide was stored at -70 ' C . Monoclonal antibody/CDllb I-domain fusion protein complexes were formed by incubating the cell lysate prepared as described above with a monoclonal antibody that recognized the FLAG peptide of the CDllb I-domain fusion protein (anti-FLAG Ml; International Biotechnologies, Inc.). These complexes were incubated with ^,25 I-rNIF and precipitated with protein A-sepharose (Calbiochem) . One microgram of Ml monoclonal antibody was incubated with 390 μl of the cell lysate that contained CDllb I-domain fusion peptide, in the presence of 10 ^s cpm ¹²⁵ I-rNIF (specific activity 47.5 μCi/μg; see Example 14(B)). The total reaction volume was 400 μl, and the reaction proceeded for 18 hours at 4'C. Immune complexes were precipitated with 100 μl protein A-Sepharose (Pharmacia) . The protein A-Sepharose was prepared as a 1:1 slurry in TACTS 20 buffer and preblocked with 0.5% BSA. After precipitation, Sepharose beads were washed three times with TACTS 20

buffer, and immune complexes were released by the addition of Tris-glycine sample buffer (Novex) under reducing conditions (100 mM dithiothreitol) . Precipitated, labeled proteins were resolved by 4-20% gradient SDS-PAGE (Novex) and visualized by autoradiography.

The ¹²⁵ I-rNIF was precipitated by Ml monoclonal antibody in the presence of cell lysate that contained recombinant I-domain peptide. In contrast, ^I23 I-rNIF was not precipitated in the absence of this cell lysate or when the Ml monoclonal antibody was substituted with a monoclonal antibody that did not react with the the FLAG portion of the fusion protein (anti-gpllbllla) . Moreover, Ml did not precipitate ^I2S I-rNIF in the presence of another fusion protein comprising FLAG and bacterial alkaline phosphatase (pFLAG-BAP; International Biotechnologies, Inc.). These data demonstrate direct., and specific interaction between NIF and the I-domain peptide of CDllb/CDlδ.